{
  "text": "大家好,今天咱们聊一个实用工具,它叫AudioCPP\n传统的本地语音模型环境,配置起来很麻烦\n庞大的音频库,复杂的Conda,还有Python,E-Lite这些痛点\n是不是很熟悉,就像Alama简化了大语言模型一样\nAudioCPC解决了音频AI的部署难题\n它是纯C++实现的原生二进制程序\n开箱挤用非常轻量,APU版本仅8兆大小\n他支持TTS声音克隆\nHello, this is audio.cpp speaking from a webpage\n支持ASR语音转文字\n支持批量合成语音\n还有实时语音功能\n我可以帮你回答问题\n翻译语言,撰写文本\n以及提供创意建议\n并且自带命令行CLI和网页界面Web UI\n咱们先看命令行工具\n也就是CLI功能\n举个例子\n输入AudioCppClare后面跟多个参数\nTask TTS表示任务是声音克隆\nFamily Pocket TTS指定模型为Pocket TTS\nModel后面写模型文件路径\nBackend CPU表示用CPU推理\n能使用GPU推理就写Cuda\nText是你要合成的文字内容\nVoice ID是参考音色文件\nOut指定输出音频的位置\n现在我们来实际操作一下\n打开AudioX CPT文件夹\n在地址栏里输入CMD回车\n这样就打开了命令提示符\n复制准备好的命令行\n粘贴进去按回车\n模型开始生成音频\n请稍等\n生成完毕\n我们打开out文件夹\n双击这个音频文件听一下\n效果还不错\n对吧\n再看第二个例子\n基本参数和前面一样\n这次使用的模型是\nQuantTTS\nVoice Reef是参考音频的位置\nReference Text是对应的参考文本\n复制这条命令切换到命令行窗口\n拦贴回车\n等待模型生成音频文件\n好,打开Out文件夹\n双击播放,听一下\n你好,Audio CPP正在Windows CPU上运行\n效果依然很棒\n说完了Cli,咱们看看Web UI\n有了命令行基础\n网页操作就更简单了\n首先,在模型列表里选择Pocket TTS\n点击加载模型\n然后选择一个参考音色文件\n点击播放视听\n在语音列表里选择英文\n点击生成语音等待模型处理\n生成完成\n我们点击播放\n听一下效果不错吧\n点击向下的小箭头\n可以下载音频文件\n接着测试Q1-3TT-S模型\n在模型列表里选择它\n这次选一个中文参考音色\n点击播放\n以前我对这句话一致半解\n现在好像有点懂了\n因为你我开始留意很多\n以前不曾关心的事\n开始对这个世界有了更多的好奇和善意\n在合成文本输入框里\n填入你想说的话\n语音列表里记得选中文\n别忘记点击加载模型\n然后点击生成语音\n好 生成完成\n点击播放听一下效果\n千万3TTS覆盖十种主要语言中文英语\n日语 韩语 德语 法语 俄语 葡萄牙语\n西班牙语和意大利语\n并提供多种方言语音配置\n以满足全球应用需求\n声音很自然\n值得注意的是合成的文本不宜过长\n不要超过100个字\n生成语音速度会很慢\n而且文本越长\n合成音频的质量越差\n如果有长文本需要处理怎么办\nAudio CPP支持Batch P处理\n我们来看一个P处理的例子\n基本参数和前面一样\nBatch Text File表示按行P处理\n把文本拆分成多行\n每行作为一段独立合成\nBatch Merge Audio Concrete\n会自动将合成好的小段音频\n合成为同一个完整文件\n看看合成文本的内容\n每一行是一个句子\n布置这条命令\n切换到命令型窗口\n粘贴回车\n等待模型生成音频文件\n好,打开AUT文件夹,双击播放,听一下\n好,咱们快速回顾一下\nAudio CPP让本地语音AI部署变得非常简单\nCLI和Web UI都支持\n涵盖TTS特隆ASR转写音乐生成和实时对话\n提供了统一的语音服务接口\n安装也只需要下载模型文件即可\n感兴趣的话,快去试试吧\n效果不错\n接下来是语音转写功能\n也就是ASR\n语音转文字\n在模型列表里选择Qin3ASR模型\n点击加载\n上传一段音频文件\n点击播放预览\n我们浏览手机端页面\nGLM5.2不具备视觉图像输入能力\n它不能给网页添加图片\n然后点击开始转写\n模型开始处理\n稍等一会儿\n转写结果出现在文本框里\n正确率非常高\n再看音乐生成选项卡\n玩法和其他语音模型类似\n选择模型加载模型\n如果模型没安装\n点击下载即可\n上传参考音频\n输入提示词\n最后点击生成音乐\n这里我不做演示了\n感兴趣的话可以自行尝试\nAudio CPP可以作为服务提供方\n共第三方应用调用\n下面聊实时语音\n它的工作原理是这样的\n麦克风输入声音\nASR模型转成文字\n文字交给大语言模型处理\n处理结果再通过TTS模型合成语音输出\n实时语音比较消耗资源\n因为后台同时加载了TTS和ASR两个模型\n要使用这个功能\n先双击RunWebBit开启WebUI服务\n然后加载一个TTS模型\n接着双击ASRBit加载ASR模型\n最后双击RealtelBit打开实时语音页面\n在设置里你可以配置LLM的接口地址\n模型名称和API Key\n点击对话\n说一声你好\n就可以看到实时交互效果了\n你好\n请问有什么\n我可以帮你了吗\n我可以帮你回答问题\n翻译语言\n撰写文本\n以及提供创意建议\n最后咱们说说怎么安装Audio CPP\n打开它的GitHub页面\n点击Releases\n下载CPU版本的压缩包\n解压后里面没有音频模型\n我们需要自己创建一个Models玩件夹\n用来存放所有模型文件\n先安装Pocket TTS\n打开Hugging Face上的模型页面\n它只有1亿参数\n非常轻量\n不用GPU也能流畅运行\n下载英文相关的模型文件\n分磁器和参考音色\n在Models下\n创建Pocket TS文件夹\n把文件放进去\n再安装QVN3 TTS\n打开摩塔社区\n搜索QVN3 TTS\n点击下载模型\n推荐用GitLFS工具来下载\n它是管理大文件的扩展\n复制GitClone地址\n在模型文件夹里打开CMD\n粘贴并回车\n如果已经下载过\n会提示已存在\n你也可以从Hugging Face的科问仓库下载\n对于不想折腾的小伙伴\n官方也提供了整合包\n包含CPU、GPU和Web UI版本\n常用音频模型都集成好了\n直接用CPU就能跑\n好,咱们快速回顾一下\nAudio CPP让本地语音AI部署变得非常简单\nCLI和Web UI都支持涵盖TTS、特隆、ASR转写音乐生成和实时对话\n提供了统一的语音服务接口\n安装也只需要下载模型文件即可\n感兴趣的话快去试试吧\n感谢观看\n我们下期再见",
  "language": "zh",
  "duration": 645.445063,
  "segments": [
    {
      "start": 0.6,
      "end": 4.46,
      "text": "大家好,今天咱们聊一个实用工具,它叫AudioCPP",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12330833023893618,
      "no_speech_prob": 1.8997281178712022e-11,
      "compression_ratio": 1.1223241590214068
    },
    {
      "start": 4.46,
      "end": 9.3,
      "text": "传统的本地语音模型环境,配置起来很麻烦",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12330833023893618,
      "no_speech_prob": 1.8997281178712022e-11,
      "compression_ratio": 1.1223241590214068
    },
    {
      "start": 9.3,
      "end": 13.56,
      "text": "庞大的音频库,复杂的Conda,还有Python,E-Lite这些痛点",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12330833023893618,
      "no_speech_prob": 1.8997281178712022e-11,
      "compression_ratio": 1.1223241590214068
    },
    {
      "start": 13.56,
      "end": 18.0,
      "text": "是不是很熟悉,就像Alama简化了大语言模型一样",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12330833023893618,
      "no_speech_prob": 1.8997281178712022e-11,
      "compression_ratio": 1.1223241590214068
    },
    {
      "start": 18.0,
      "end": 21.46,
      "text": "AudioCPC解决了音频AI的部署难题",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12330833023893618,
      "no_speech_prob": 1.8997281178712022e-11,
      "compression_ratio": 1.1223241590214068
    },
    {
      "start": 21.46,
      "end": 24.88,
      "text": "它是纯C++实现的原生二进制程序",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12330833023893618,
      "no_speech_prob": 1.8997281178712022e-11,
      "compression_ratio": 1.1223241590214068
    },
    {
      "start": 24.88,
      "end": 29.38,
      "text": "开箱挤用非常轻量,APU版本仅8兆大小",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12330833023893618,
      "no_speech_prob": 1.8997281178712022e-11,
      "compression_ratio": 1.1223241590214068
    },
    {
      "start": 29.38,
      "end": 31.66,
      "text": "他支持TTS声音克隆",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.18088975886708683,
      "no_speech_prob": 8.905373834189323e-12,
      "compression_ratio": 1.0185185185185186
    },
    {
      "start": 31.66,
      "end": 34.84,
      "text": "Hello, this is audio.cpp speaking from a webpage",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.18088975886708683,
      "no_speech_prob": 8.905373834189323e-12,
      "compression_ratio": 1.0185185185185186
    },
    {
      "start": 34.84,
      "end": 37.1,
      "text": "支持ASR语音转文字",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.18088975886708683,
      "no_speech_prob": 8.905373834189323e-12,
      "compression_ratio": 1.0185185185185186
    },
    {
      "start": 37.1,
      "end": 44.38,
      "text": "支持批量合成语音",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.18088975886708683,
      "no_speech_prob": 8.905373834189323e-12,
      "compression_ratio": 1.0185185185185186
    },
    {
      "start": 44.38,
      "end": 50.94,
      "text": "还有实时语音功能",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.18088975886708683,
      "no_speech_prob": 8.905373834189323e-12,
      "compression_ratio": 1.0185185185185186
    },
    {
      "start": 50.94,
      "end": 53.94,
      "text": "我可以帮你回答问题",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.18088975886708683,
      "no_speech_prob": 8.905373834189323e-12,
      "compression_ratio": 1.0185185185185186
    },
    {
      "start": 53.94,
      "end": 55.72,
      "text": "翻译语言,撰写文本",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.18088975886708683,
      "no_speech_prob": 8.905373834189323e-12,
      "compression_ratio": 1.0185185185185186
    },
    {
      "start": 55.72,
      "end": 57.18,
      "text": "以及提供创意建议",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.18088975886708683,
      "no_speech_prob": 8.905373834189323e-12,
      "compression_ratio": 1.0185185185185186
    },
    {
      "start": 57.18,
      "end": 62.92,
      "text": "并且自带命令行CLI和网页界面Web UI",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.16451344247591698,
      "no_speech_prob": 1.1257347484749936e-11,
      "compression_ratio": 1.0726495726495726
    },
    {
      "start": 62.92,
      "end": 65.74,
      "text": "咱们先看命令行工具",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.16451344247591698,
      "no_speech_prob": 1.1257347484749936e-11,
      "compression_ratio": 1.0726495726495726
    },
    {
      "start": 65.74,
      "end": 68.42,
      "text": "也就是CLI功能",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.16451344247591698,
      "no_speech_prob": 1.1257347484749936e-11,
      "compression_ratio": 1.0726495726495726
    },
    {
      "start": 68.42,
      "end": 70.36,
      "text": "举个例子",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.16451344247591698,
      "no_speech_prob": 1.1257347484749936e-11,
      "compression_ratio": 1.0726495726495726
    },
    {
      "start": 70.36,
      "end": 74.82,
      "text": "输入AudioCppClare后面跟多个参数",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.16451344247591698,
      "no_speech_prob": 1.1257347484749936e-11,
      "compression_ratio": 1.0726495726495726
    },
    {
      "start": 74.82,
      "end": 78.62,
      "text": "Task TTS表示任务是声音克隆",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.16451344247591698,
      "no_speech_prob": 1.1257347484749936e-11,
      "compression_ratio": 1.0726495726495726
    },
    {
      "start": 78.62,
      "end": 82.9,
      "text": "Family Pocket TTS指定模型为Pocket TTS",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.16451344247591698,
      "no_speech_prob": 1.1257347484749936e-11,
      "compression_ratio": 1.0726495726495726
    },
    {
      "start": 82.9,
      "end": 86.76,
      "text": "Model后面写模型文件路径",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.16451344247591698,
      "no_speech_prob": 1.1257347484749936e-11,
      "compression_ratio": 1.0726495726495726
    },
    {
      "start": 86.76,
      "end": 89.88,
      "text": "Backend CPU表示用CPU推理",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1389306430158944,
      "no_speech_prob": 1.2124220030718114e-11,
      "compression_ratio": 1.1232394366197183
    },
    {
      "start": 89.88,
      "end": 92.78,
      "text": "能使用GPU推理就写Cuda",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1389306430158944,
      "no_speech_prob": 1.2124220030718114e-11,
      "compression_ratio": 1.1232394366197183
    },
    {
      "start": 92.78,
      "end": 95.68,
      "text": "Text是你要合成的文字内容",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1389306430158944,
      "no_speech_prob": 1.2124220030718114e-11,
      "compression_ratio": 1.1232394366197183
    },
    {
      "start": 95.68,
      "end": 98.46,
      "text": "Voice ID是参考音色文件",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1389306430158944,
      "no_speech_prob": 1.2124220030718114e-11,
      "compression_ratio": 1.1232394366197183
    },
    {
      "start": 98.46,
      "end": 101.64,
      "text": "Out指定输出音频的位置",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1389306430158944,
      "no_speech_prob": 1.2124220030718114e-11,
      "compression_ratio": 1.1232394366197183
    },
    {
      "start": 101.64,
      "end": 104.28,
      "text": "现在我们来实际操作一下",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1389306430158944,
      "no_speech_prob": 1.2124220030718114e-11,
      "compression_ratio": 1.1232394366197183
    },
    {
      "start": 104.28,
      "end": 106.56,
      "text": "打开AudioX CPT文件夹",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1389306430158944,
      "no_speech_prob": 1.2124220030718114e-11,
      "compression_ratio": 1.1232394366197183
    },
    {
      "start": 106.56,
      "end": 109.24,
      "text": "在地址栏里输入CMD回车",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1389306430158944,
      "no_speech_prob": 1.2124220030718114e-11,
      "compression_ratio": 1.1232394366197183
    },
    {
      "start": 109.24,
      "end": 111.64,
      "text": "这样就打开了命令提示符",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1389306430158944,
      "no_speech_prob": 1.2124220030718114e-11,
      "compression_ratio": 1.1232394366197183
    },
    {
      "start": 111.64,
      "end": 113.12,
      "text": "复制准备好的命令行",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1389306430158944,
      "no_speech_prob": 1.2124220030718114e-11,
      "compression_ratio": 1.1232394366197183
    },
    {
      "start": 113.12,
      "end": 115.8,
      "text": "粘贴进去按回车",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1389306430158944,
      "no_speech_prob": 1.2124220030718114e-11,
      "compression_ratio": 1.1232394366197183
    },
    {
      "start": 115.8,
      "end": 118.38,
      "text": "模型开始生成音频",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.24351405071955856,
      "no_speech_prob": 1.0265334589310005e-11,
      "compression_ratio": 1.1092896174863387
    },
    {
      "start": 118.38,
      "end": 119.54,
      "text": "请稍等",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.24351405071955856,
      "no_speech_prob": 1.0265334589310005e-11,
      "compression_ratio": 1.1092896174863387
    },
    {
      "start": 119.54,
      "end": 120.52,
      "text": "生成完毕",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.24351405071955856,
      "no_speech_prob": 1.0265334589310005e-11,
      "compression_ratio": 1.1092896174863387
    },
    {
      "start": 120.52,
      "end": 122.7,
      "text": "我们打开out文件夹",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.24351405071955856,
      "no_speech_prob": 1.0265334589310005e-11,
      "compression_ratio": 1.1092896174863387
    },
    {
      "start": 122.7,
      "end": 125.36,
      "text": "双击这个音频文件听一下",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.24351405071955856,
      "no_speech_prob": 1.0265334589310005e-11,
      "compression_ratio": 1.1092896174863387
    },
    {
      "start": 134.48,
      "end": 136.44,
      "text": "效果还不错",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.24351405071955856,
      "no_speech_prob": 1.0265334589310005e-11,
      "compression_ratio": 1.1092896174863387
    },
    {
      "start": 136.44,
      "end": 137.0,
      "text": "对吧",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.24351405071955856,
      "no_speech_prob": 1.0265334589310005e-11,
      "compression_ratio": 1.1092896174863387
    },
    {
      "start": 137.0,
      "end": 139.12,
      "text": "再看第二个例子",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.24351405071955856,
      "no_speech_prob": 1.0265334589310005e-11,
      "compression_ratio": 1.1092896174863387
    },
    {
      "start": 139.12,
      "end": 141.28,
      "text": "基本参数和前面一样",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.24351405071955856,
      "no_speech_prob": 1.0265334589310005e-11,
      "compression_ratio": 1.1092896174863387
    },
    {
      "start": 141.28,
      "end": 142.84,
      "text": "这次使用的模型是",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.24351405071955856,
      "no_speech_prob": 1.0265334589310005e-11,
      "compression_ratio": 1.1092896174863387
    },
    {
      "start": 142.84,
      "end": 143.96,
      "text": "QuantTTS",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.24351405071955856,
      "no_speech_prob": 1.0265334589310005e-11,
      "compression_ratio": 1.1092896174863387
    },
    {
      "start": 143.96,
      "end": 148.58,
      "text": "Voice Reef是参考音频的位置",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20429946564056062,
      "no_speech_prob": 8.102263651665886e-12,
      "compression_ratio": 1.0526315789473684
    },
    {
      "start": 148.58,
      "end": 151.38,
      "text": "Reference Text是对应的参考文本",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20429946564056062,
      "no_speech_prob": 8.102263651665886e-12,
      "compression_ratio": 1.0526315789473684
    },
    {
      "start": 151.38,
      "end": 155.64,
      "text": "复制这条命令切换到命令行窗口",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20429946564056062,
      "no_speech_prob": 8.102263651665886e-12,
      "compression_ratio": 1.0526315789473684
    },
    {
      "start": 155.64,
      "end": 157.16,
      "text": "拦贴回车",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20429946564056062,
      "no_speech_prob": 8.102263651665886e-12,
      "compression_ratio": 1.0526315789473684
    },
    {
      "start": 157.16,
      "end": 159.96,
      "text": "等待模型生成音频文件",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20429946564056062,
      "no_speech_prob": 8.102263651665886e-12,
      "compression_ratio": 1.0526315789473684
    },
    {
      "start": 159.96,
      "end": 167.04,
      "text": "好,打开Out文件夹",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20429946564056062,
      "no_speech_prob": 8.102263651665886e-12,
      "compression_ratio": 1.0526315789473684
    },
    {
      "start": 167.04,
      "end": 170.28,
      "text": "双击播放,听一下",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20429946564056062,
      "no_speech_prob": 8.102263651665886e-12,
      "compression_ratio": 1.0526315789473684
    },
    {
      "start": 173.96,
      "end": 178.52,
      "text": "你好,Audio CPP正在Windows CPU上运行",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1088802994800215,
      "no_speech_prob": 1.99131267031305e-11,
      "compression_ratio": 1.0327868852459017
    },
    {
      "start": 178.52,
      "end": 180.44,
      "text": "效果依然很棒",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1088802994800215,
      "no_speech_prob": 1.99131267031305e-11,
      "compression_ratio": 1.0327868852459017
    },
    {
      "start": 180.44,
      "end": 183.74,
      "text": "说完了Cli,咱们看看Web UI",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1088802994800215,
      "no_speech_prob": 1.99131267031305e-11,
      "compression_ratio": 1.0327868852459017
    },
    {
      "start": 183.74,
      "end": 185.64,
      "text": "有了命令行基础",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1088802994800215,
      "no_speech_prob": 1.99131267031305e-11,
      "compression_ratio": 1.0327868852459017
    },
    {
      "start": 185.64,
      "end": 187.4,
      "text": "网页操作就更简单了",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1088802994800215,
      "no_speech_prob": 1.99131267031305e-11,
      "compression_ratio": 1.0327868852459017
    },
    {
      "start": 187.4,
      "end": 191.14,
      "text": "首先,在模型列表里选择Pocket TTS",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1088802994800215,
      "no_speech_prob": 1.99131267031305e-11,
      "compression_ratio": 1.0327868852459017
    },
    {
      "start": 191.14,
      "end": 192.56,
      "text": "点击加载模型",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1088802994800215,
      "no_speech_prob": 1.99131267031305e-11,
      "compression_ratio": 1.0327868852459017
    },
    {
      "start": 192.56,
      "end": 194.64,
      "text": "然后选择一个参考音色文件",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1088802994800215,
      "no_speech_prob": 1.99131267031305e-11,
      "compression_ratio": 1.0327868852459017
    },
    {
      "start": 194.64,
      "end": 197.84,
      "text": "点击播放视听",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1088802994800215,
      "no_speech_prob": 1.99131267031305e-11,
      "compression_ratio": 1.0327868852459017
    },
    {
      "start": 197.84,
      "end": 208.66,
      "text": "在语音列表里选择英文",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1697694554048426,
      "no_speech_prob": 7.325091921916993e-12,
      "compression_ratio": 1.1162790697674418
    },
    {
      "start": 208.66,
      "end": 210.72,
      "text": "点击生成语音等待模型处理",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1697694554048426,
      "no_speech_prob": 7.325091921916993e-12,
      "compression_ratio": 1.1162790697674418
    },
    {
      "start": 218.32,
      "end": 219.14,
      "text": "生成完成",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1697694554048426,
      "no_speech_prob": 7.325091921916993e-12,
      "compression_ratio": 1.1162790697674418
    },
    {
      "start": 219.14,
      "end": 220.1,
      "text": "我们点击播放",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1697694554048426,
      "no_speech_prob": 7.325091921916993e-12,
      "compression_ratio": 1.1162790697674418
    },
    {
      "start": 220.1,
      "end": 221.82,
      "text": "听一下效果不错吧",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1697694554048426,
      "no_speech_prob": 7.325091921916993e-12,
      "compression_ratio": 1.1162790697674418
    },
    {
      "start": 221.82,
      "end": 226.58,
      "text": "点击向下的小箭头",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1697694554048426,
      "no_speech_prob": 7.325091921916993e-12,
      "compression_ratio": 1.1162790697674418
    },
    {
      "start": 226.58,
      "end": 228.36,
      "text": "可以下载音频文件",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13716577802385602,
      "no_speech_prob": 1.2612629690655908e-11,
      "compression_ratio": 1.2481203007518797
    },
    {
      "start": 228.36,
      "end": 231.1,
      "text": "接着测试Q1-3TT-S模型",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13716577802385602,
      "no_speech_prob": 1.2612629690655908e-11,
      "compression_ratio": 1.2481203007518797
    },
    {
      "start": 231.1,
      "end": 233.36,
      "text": "在模型列表里选择它",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13716577802385602,
      "no_speech_prob": 1.2612629690655908e-11,
      "compression_ratio": 1.2481203007518797
    },
    {
      "start": 233.36,
      "end": 239.3,
      "text": "这次选一个中文参考音色",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13716577802385602,
      "no_speech_prob": 1.2612629690655908e-11,
      "compression_ratio": 1.2481203007518797
    },
    {
      "start": 239.3,
      "end": 240.34,
      "text": "点击播放",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13716577802385602,
      "no_speech_prob": 1.2612629690655908e-11,
      "compression_ratio": 1.2481203007518797
    },
    {
      "start": 240.34,
      "end": 243.52,
      "text": "以前我对这句话一致半解",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13716577802385602,
      "no_speech_prob": 1.2612629690655908e-11,
      "compression_ratio": 1.2481203007518797
    },
    {
      "start": 243.52,
      "end": 245.08,
      "text": "现在好像有点懂了",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13716577802385602,
      "no_speech_prob": 1.2612629690655908e-11,
      "compression_ratio": 1.2481203007518797
    },
    {
      "start": 245.08,
      "end": 247.02,
      "text": "因为你我开始留意很多",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13716577802385602,
      "no_speech_prob": 1.2612629690655908e-11,
      "compression_ratio": 1.2481203007518797
    },
    {
      "start": 247.02,
      "end": 248.46,
      "text": "以前不曾关心的事",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13716577802385602,
      "no_speech_prob": 1.2612629690655908e-11,
      "compression_ratio": 1.2481203007518797
    },
    {
      "start": 248.46,
      "end": 251.58,
      "text": "开始对这个世界有了更多的好奇和善意",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13716577802385602,
      "no_speech_prob": 1.2612629690655908e-11,
      "compression_ratio": 1.2481203007518797
    },
    {
      "start": 251.58,
      "end": 254.02,
      "text": "在合成文本输入框里",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13716577802385602,
      "no_speech_prob": 1.2612629690655908e-11,
      "compression_ratio": 1.2481203007518797
    },
    {
      "start": 254.02,
      "end": 255.22,
      "text": "填入你想说的话",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13716577802385602,
      "no_speech_prob": 1.2612629690655908e-11,
      "compression_ratio": 1.2481203007518797
    },
    {
      "start": 255.22,
      "end": 257.28,
      "text": "语音列表里记得选中文",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11318788836079259,
      "no_speech_prob": 1.1385562631582857e-11,
      "compression_ratio": 1.2666666666666666
    },
    {
      "start": 257.28,
      "end": 258.88,
      "text": "别忘记点击加载模型",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11318788836079259,
      "no_speech_prob": 1.1385562631582857e-11,
      "compression_ratio": 1.2666666666666666
    },
    {
      "start": 258.88,
      "end": 262.52,
      "text": "然后点击生成语音",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11318788836079259,
      "no_speech_prob": 1.1385562631582857e-11,
      "compression_ratio": 1.2666666666666666
    },
    {
      "start": 262.52,
      "end": 268.08,
      "text": "好 生成完成",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11318788836079259,
      "no_speech_prob": 1.1385562631582857e-11,
      "compression_ratio": 1.2666666666666666
    },
    {
      "start": 268.08,
      "end": 270.32,
      "text": "点击播放听一下效果",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11318788836079259,
      "no_speech_prob": 1.1385562631582857e-11,
      "compression_ratio": 1.2666666666666666
    },
    {
      "start": 270.32,
      "end": 274.02,
      "text": "千万3TTS覆盖十种主要语言中文英语",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11318788836079259,
      "no_speech_prob": 1.1385562631582857e-11,
      "compression_ratio": 1.2666666666666666
    },
    {
      "start": 274.02,
      "end": 277.66,
      "text": "日语 韩语 德语 法语 俄语 葡萄牙语",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11318788836079259,
      "no_speech_prob": 1.1385562631582857e-11,
      "compression_ratio": 1.2666666666666666
    },
    {
      "start": 277.66,
      "end": 279.3,
      "text": "西班牙语和意大利语",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11318788836079259,
      "no_speech_prob": 1.1385562631582857e-11,
      "compression_ratio": 1.2666666666666666
    },
    {
      "start": 279.3,
      "end": 281.66,
      "text": "并提供多种方言语音配置",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11318788836079259,
      "no_speech_prob": 1.1385562631582857e-11,
      "compression_ratio": 1.2666666666666666
    },
    {
      "start": 281.66,
      "end": 283.7,
      "text": "以满足全球应用需求",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11318788836079259,
      "no_speech_prob": 1.1385562631582857e-11,
      "compression_ratio": 1.2666666666666666
    },
    {
      "start": 283.7,
      "end": 285.48,
      "text": "声音很自然",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13731935456043154,
      "no_speech_prob": 9.938037372203556e-12,
      "compression_ratio": 1.1753731343283582
    },
    {
      "start": 285.48,
      "end": 291.12,
      "text": "值得注意的是合成的文本不宜过长",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13731935456043154,
      "no_speech_prob": 9.938037372203556e-12,
      "compression_ratio": 1.1753731343283582
    },
    {
      "start": 291.12,
      "end": 294.06,
      "text": "不要超过100个字",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13731935456043154,
      "no_speech_prob": 9.938037372203556e-12,
      "compression_ratio": 1.1753731343283582
    },
    {
      "start": 294.06,
      "end": 295.6,
      "text": "生成语音速度会很慢",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13731935456043154,
      "no_speech_prob": 9.938037372203556e-12,
      "compression_ratio": 1.1753731343283582
    },
    {
      "start": 295.6,
      "end": 296.76,
      "text": "而且文本越长",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13731935456043154,
      "no_speech_prob": 9.938037372203556e-12,
      "compression_ratio": 1.1753731343283582
    },
    {
      "start": 296.76,
      "end": 298.42,
      "text": "合成音频的质量越差",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13731935456043154,
      "no_speech_prob": 9.938037372203556e-12,
      "compression_ratio": 1.1753731343283582
    },
    {
      "start": 298.42,
      "end": 300.56,
      "text": "如果有长文本需要处理怎么办",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13731935456043154,
      "no_speech_prob": 9.938037372203556e-12,
      "compression_ratio": 1.1753731343283582
    },
    {
      "start": 300.56,
      "end": 305.04,
      "text": "Audio CPP支持Batch P处理",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13731935456043154,
      "no_speech_prob": 9.938037372203556e-12,
      "compression_ratio": 1.1753731343283582
    },
    {
      "start": 305.04,
      "end": 307.6,
      "text": "我们来看一个P处理的例子",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13731935456043154,
      "no_speech_prob": 9.938037372203556e-12,
      "compression_ratio": 1.1753731343283582
    },
    {
      "start": 307.6,
      "end": 309.58,
      "text": "基本参数和前面一样",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13731935456043154,
      "no_speech_prob": 9.938037372203556e-12,
      "compression_ratio": 1.1753731343283582
    },
    {
      "start": 309.58,
      "end": 312.52,
      "text": "Batch Text File表示按行P处理",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13731935456043154,
      "no_speech_prob": 9.938037372203556e-12,
      "compression_ratio": 1.1753731343283582
    },
    {
      "start": 312.52,
      "end": 314.66,
      "text": "把文本拆分成多行",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11354912741709564,
      "no_speech_prob": 1.2785500089207424e-11,
      "compression_ratio": 1.180672268907563
    },
    {
      "start": 314.66,
      "end": 316.72,
      "text": "每行作为一段独立合成",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11354912741709564,
      "no_speech_prob": 1.2785500089207424e-11,
      "compression_ratio": 1.180672268907563
    },
    {
      "start": 316.72,
      "end": 319.68,
      "text": "Batch Merge Audio Concrete",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11354912741709564,
      "no_speech_prob": 1.2785500089207424e-11,
      "compression_ratio": 1.180672268907563
    },
    {
      "start": 319.68,
      "end": 321.88,
      "text": "会自动将合成好的小段音频",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11354912741709564,
      "no_speech_prob": 1.2785500089207424e-11,
      "compression_ratio": 1.180672268907563
    },
    {
      "start": 321.88,
      "end": 323.78,
      "text": "合成为同一个完整文件",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11354912741709564,
      "no_speech_prob": 1.2785500089207424e-11,
      "compression_ratio": 1.180672268907563
    },
    {
      "start": 323.78,
      "end": 325.46,
      "text": "看看合成文本的内容",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11354912741709564,
      "no_speech_prob": 1.2785500089207424e-11,
      "compression_ratio": 1.180672268907563
    },
    {
      "start": 325.46,
      "end": 327.2,
      "text": "每一行是一个句子",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11354912741709564,
      "no_speech_prob": 1.2785500089207424e-11,
      "compression_ratio": 1.180672268907563
    },
    {
      "start": 327.2,
      "end": 328.42,
      "text": "布置这条命令",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11354912741709564,
      "no_speech_prob": 1.2785500089207424e-11,
      "compression_ratio": 1.180672268907563
    },
    {
      "start": 328.42,
      "end": 335.06,
      "text": "切换到命令型窗口",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11354912741709564,
      "no_speech_prob": 1.2785500089207424e-11,
      "compression_ratio": 1.180672268907563
    },
    {
      "start": 335.06,
      "end": 336.44,
      "text": "粘贴回车",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11354912741709564,
      "no_speech_prob": 1.2785500089207424e-11,
      "compression_ratio": 1.180672268907563
    },
    {
      "start": 336.44,
      "end": 339.46,
      "text": "等待模型生成音频文件",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11354912741709564,
      "no_speech_prob": 1.2785500089207424e-11,
      "compression_ratio": 1.180672268907563
    },
    {
      "start": 342.52,
      "end": 346.54,
      "text": "好,打开AUT文件夹,双击播放,听一下",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10937793382251536,
      "no_speech_prob": 1.5439757111712638e-11,
      "compression_ratio": 1.0790513833992095
    },
    {
      "start": 346.54,
      "end": 352.6,
      "text": "好,咱们快速回顾一下",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10937793382251536,
      "no_speech_prob": 1.5439757111712638e-11,
      "compression_ratio": 1.0790513833992095
    },
    {
      "start": 352.6,
      "end": 356.58,
      "text": "Audio CPP让本地语音AI部署变得非常简单",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10937793382251536,
      "no_speech_prob": 1.5439757111712638e-11,
      "compression_ratio": 1.0790513833992095
    },
    {
      "start": 356.58,
      "end": 359.36,
      "text": "CLI和Web UI都支持",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10937793382251536,
      "no_speech_prob": 1.5439757111712638e-11,
      "compression_ratio": 1.0790513833992095
    },
    {
      "start": 359.36,
      "end": 364.42,
      "text": "涵盖TTS特隆ASR转写音乐生成和实时对话",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10937793382251536,
      "no_speech_prob": 1.5439757111712638e-11,
      "compression_ratio": 1.0790513833992095
    },
    {
      "start": 364.42,
      "end": 367.2,
      "text": "提供了统一的语音服务接口",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10937793382251536,
      "no_speech_prob": 1.5439757111712638e-11,
      "compression_ratio": 1.0790513833992095
    },
    {
      "start": 367.2,
      "end": 370.78,
      "text": "安装也只需要下载模型文件即可",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10937793382251536,
      "no_speech_prob": 1.5439757111712638e-11,
      "compression_ratio": 1.0790513833992095
    },
    {
      "start": 370.78,
      "end": 372.78,
      "text": "感兴趣的话,快去试试吧",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12088712056477864,
      "no_speech_prob": 1.5147513451885253e-11,
      "compression_ratio": 1.187725631768953
    },
    {
      "start": 372.78,
      "end": 374.1,
      "text": "效果不错",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12088712056477864,
      "no_speech_prob": 1.5147513451885253e-11,
      "compression_ratio": 1.187725631768953
    },
    {
      "start": 374.1,
      "end": 376.62,
      "text": "接下来是语音转写功能",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12088712056477864,
      "no_speech_prob": 1.5147513451885253e-11,
      "compression_ratio": 1.187725631768953
    },
    {
      "start": 376.62,
      "end": 377.94,
      "text": "也就是ASR",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12088712056477864,
      "no_speech_prob": 1.5147513451885253e-11,
      "compression_ratio": 1.187725631768953
    },
    {
      "start": 377.94,
      "end": 379.4,
      "text": "语音转文字",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12088712056477864,
      "no_speech_prob": 1.5147513451885253e-11,
      "compression_ratio": 1.187725631768953
    },
    {
      "start": 379.4,
      "end": 383.4,
      "text": "在模型列表里选择Qin3ASR模型",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12088712056477864,
      "no_speech_prob": 1.5147513451885253e-11,
      "compression_ratio": 1.187725631768953
    },
    {
      "start": 383.4,
      "end": 384.46,
      "text": "点击加载",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12088712056477864,
      "no_speech_prob": 1.5147513451885253e-11,
      "compression_ratio": 1.187725631768953
    },
    {
      "start": 384.46,
      "end": 386.38,
      "text": "上传一段音频文件",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12088712056477864,
      "no_speech_prob": 1.5147513451885253e-11,
      "compression_ratio": 1.187725631768953
    },
    {
      "start": 386.38,
      "end": 387.58,
      "text": "点击播放预览",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12088712056477864,
      "no_speech_prob": 1.5147513451885253e-11,
      "compression_ratio": 1.187725631768953
    },
    {
      "start": 387.58,
      "end": 389.74,
      "text": "我们浏览手机端页面",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12088712056477864,
      "no_speech_prob": 1.5147513451885253e-11,
      "compression_ratio": 1.187725631768953
    },
    {
      "start": 389.74,
      "end": 394.34,
      "text": "GLM5.2不具备视觉图像输入能力",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12088712056477864,
      "no_speech_prob": 1.5147513451885253e-11,
      "compression_ratio": 1.187725631768953
    },
    {
      "start": 394.34,
      "end": 396.24,
      "text": "它不能给网页添加图片",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12088712056477864,
      "no_speech_prob": 1.5147513451885253e-11,
      "compression_ratio": 1.187725631768953
    },
    {
      "start": 396.24,
      "end": 398.78,
      "text": "然后点击开始转写",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12088712056477864,
      "no_speech_prob": 1.5147513451885253e-11,
      "compression_ratio": 1.187725631768953
    },
    {
      "start": 398.78,
      "end": 399.9,
      "text": "模型开始处理",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12088712056477864,
      "no_speech_prob": 1.5147513451885253e-11,
      "compression_ratio": 1.187725631768953
    },
    {
      "start": 399.9,
      "end": 400.84,
      "text": "稍等一会儿",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12078539965903923,
      "no_speech_prob": 1.1119177627250121e-11,
      "compression_ratio": 1.2469135802469136
    },
    {
      "start": 400.84,
      "end": 402.72,
      "text": "转写结果出现在文本框里",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12078539965903923,
      "no_speech_prob": 1.1119177627250121e-11,
      "compression_ratio": 1.2469135802469136
    },
    {
      "start": 402.72,
      "end": 404.0,
      "text": "正确率非常高",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12078539965903923,
      "no_speech_prob": 1.1119177627250121e-11,
      "compression_ratio": 1.2469135802469136
    },
    {
      "start": 404.0,
      "end": 409.04,
      "text": "再看音乐生成选项卡",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12078539965903923,
      "no_speech_prob": 1.1119177627250121e-11,
      "compression_ratio": 1.2469135802469136
    },
    {
      "start": 409.04,
      "end": 411.86,
      "text": "玩法和其他语音模型类似",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12078539965903923,
      "no_speech_prob": 1.1119177627250121e-11,
      "compression_ratio": 1.2469135802469136
    },
    {
      "start": 411.86,
      "end": 414.14,
      "text": "选择模型加载模型",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12078539965903923,
      "no_speech_prob": 1.1119177627250121e-11,
      "compression_ratio": 1.2469135802469136
    },
    {
      "start": 414.14,
      "end": 416.0,
      "text": "如果模型没安装",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12078539965903923,
      "no_speech_prob": 1.1119177627250121e-11,
      "compression_ratio": 1.2469135802469136
    },
    {
      "start": 416.0,
      "end": 417.4,
      "text": "点击下载即可",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12078539965903923,
      "no_speech_prob": 1.1119177627250121e-11,
      "compression_ratio": 1.2469135802469136
    },
    {
      "start": 417.4,
      "end": 419.78,
      "text": "上传参考音频",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12078539965903923,
      "no_speech_prob": 1.1119177627250121e-11,
      "compression_ratio": 1.2469135802469136
    },
    {
      "start": 419.78,
      "end": 420.92,
      "text": "输入提示词",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12078539965903923,
      "no_speech_prob": 1.1119177627250121e-11,
      "compression_ratio": 1.2469135802469136
    },
    {
      "start": 420.92,
      "end": 423.0,
      "text": "最后点击生成音乐",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12078539965903923,
      "no_speech_prob": 1.1119177627250121e-11,
      "compression_ratio": 1.2469135802469136
    },
    {
      "start": 423.0,
      "end": 424.92,
      "text": "这里我不做演示了",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12078539965903923,
      "no_speech_prob": 1.1119177627250121e-11,
      "compression_ratio": 1.2469135802469136
    },
    {
      "start": 424.92,
      "end": 427.12,
      "text": "感兴趣的话可以自行尝试",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12078539965903923,
      "no_speech_prob": 1.1119177627250121e-11,
      "compression_ratio": 1.2469135802469136
    },
    {
      "start": 427.12,
      "end": 430.7,
      "text": "Audio CPP可以作为服务提供方",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12215950855842003,
      "no_speech_prob": 8.805031356806658e-12,
      "compression_ratio": 1.125
    },
    {
      "start": 430.7,
      "end": 433.18,
      "text": "共第三方应用调用",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12215950855842003,
      "no_speech_prob": 8.805031356806658e-12,
      "compression_ratio": 1.125
    },
    {
      "start": 433.18,
      "end": 438.0,
      "text": "下面聊实时语音",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12215950855842003,
      "no_speech_prob": 8.805031356806658e-12,
      "compression_ratio": 1.125
    },
    {
      "start": 438.0,
      "end": 440.8,
      "text": "它的工作原理是这样的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12215950855842003,
      "no_speech_prob": 8.805031356806658e-12,
      "compression_ratio": 1.125
    },
    {
      "start": 440.8,
      "end": 442.42,
      "text": "麦克风输入声音",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12215950855842003,
      "no_speech_prob": 8.805031356806658e-12,
      "compression_ratio": 1.125
    },
    {
      "start": 442.42,
      "end": 445.3,
      "text": "ASR模型转成文字",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12215950855842003,
      "no_speech_prob": 8.805031356806658e-12,
      "compression_ratio": 1.125
    },
    {
      "start": 445.3,
      "end": 448.66,
      "text": "文字交给大语言模型处理",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12215950855842003,
      "no_speech_prob": 8.805031356806658e-12,
      "compression_ratio": 1.125
    },
    {
      "start": 448.66,
      "end": 453.42,
      "text": "处理结果再通过TTS模型合成语音输出",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12215950855842003,
      "no_speech_prob": 8.805031356806658e-12,
      "compression_ratio": 1.125
    },
    {
      "start": 453.883,
      "end": 456.303,
      "text": "实时语音比较消耗资源",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.13148527940114338,
      "no_speech_prob": 1.1128795801562674e-11,
      "compression_ratio": 1.1822660098522169
    },
    {
      "start": 456.303,
      "end": 461.323,
      "text": "因为后台同时加载了TTS和ASR两个模型",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.13148527940114338,
      "no_speech_prob": 1.1128795801562674e-11,
      "compression_ratio": 1.1822660098522169
    },
    {
      "start": 461.323,
      "end": 463.223,
      "text": "要使用这个功能",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.13148527940114338,
      "no_speech_prob": 1.1128795801562674e-11,
      "compression_ratio": 1.1822660098522169
    },
    {
      "start": 463.223,
      "end": 468.183,
      "text": "先双击RunWebBit开启WebUI服务",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.13148527940114338,
      "no_speech_prob": 1.1128795801562674e-11,
      "compression_ratio": 1.1822660098522169
    },
    {
      "start": 468.183,
      "end": 470.883,
      "text": "然后加载一个TTS模型",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.13148527940114338,
      "no_speech_prob": 1.1128795801562674e-11,
      "compression_ratio": 1.1822660098522169
    },
    {
      "start": 470.883,
      "end": 475.743,
      "text": "接着双击ASRBit加载ASR模型",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.13148527940114338,
      "no_speech_prob": 1.1128795801562674e-11,
      "compression_ratio": 1.1822660098522169
    },
    {
      "start": 475.743,
      "end": 480.623,
      "text": "最后双击RealtelBit打开实时语音页面",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.13148527940114338,
      "no_speech_prob": 1.1128795801562674e-11,
      "compression_ratio": 1.1822660098522169
    },
    {
      "start": 480.623,
      "end": 485.963,
      "text": "在设置里你可以配置LLM的接口地址",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16216283185141428,
      "no_speech_prob": 1.03344980142972e-11,
      "compression_ratio": 1.1706161137440758
    },
    {
      "start": 485.963,
      "end": 488.183,
      "text": "模型名称和API Key",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16216283185141428,
      "no_speech_prob": 1.03344980142972e-11,
      "compression_ratio": 1.1706161137440758
    },
    {
      "start": 488.183,
      "end": 489.303,
      "text": "点击对话",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16216283185141428,
      "no_speech_prob": 1.03344980142972e-11,
      "compression_ratio": 1.1706161137440758
    },
    {
      "start": 489.303,
      "end": 490.583,
      "text": "说一声你好",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16216283185141428,
      "no_speech_prob": 1.03344980142972e-11,
      "compression_ratio": 1.1706161137440758
    },
    {
      "start": 490.583,
      "end": 493.343,
      "text": "就可以看到实时交互效果了",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16216283185141428,
      "no_speech_prob": 1.03344980142972e-11,
      "compression_ratio": 1.1706161137440758
    },
    {
      "start": 493.343,
      "end": 496.823,
      "text": "你好",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16216283185141428,
      "no_speech_prob": 1.03344980142972e-11,
      "compression_ratio": 1.1706161137440758
    },
    {
      "start": 496.823,
      "end": 499.203,
      "text": "请问有什么",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16216283185141428,
      "no_speech_prob": 1.03344980142972e-11,
      "compression_ratio": 1.1706161137440758
    },
    {
      "start": 499.203,
      "end": 500.343,
      "text": "我可以帮你了吗",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16216283185141428,
      "no_speech_prob": 1.03344980142972e-11,
      "compression_ratio": 1.1706161137440758
    },
    {
      "start": 500.343,
      "end": 506.083,
      "text": "我可以帮你回答问题",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16216283185141428,
      "no_speech_prob": 1.03344980142972e-11,
      "compression_ratio": 1.1706161137440758
    },
    {
      "start": 506.083,
      "end": 506.963,
      "text": "翻译语言",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16216283185141428,
      "no_speech_prob": 1.03344980142972e-11,
      "compression_ratio": 1.1706161137440758
    },
    {
      "start": 506.963,
      "end": 507.863,
      "text": "撰写文本",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16216283185141428,
      "no_speech_prob": 1.03344980142972e-11,
      "compression_ratio": 1.1706161137440758
    },
    {
      "start": 507.863,
      "end": 509.343,
      "text": "以及提供创意建议",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16216283185141428,
      "no_speech_prob": 1.03344980142972e-11,
      "compression_ratio": 1.1706161137440758
    },
    {
      "start": 509.343,
      "end": 513.743,
      "text": "最后咱们说说怎么安装Audio CPP",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14765555171643274,
      "no_speech_prob": 9.698554459536268e-12,
      "compression_ratio": 1.072289156626506
    },
    {
      "start": 513.743,
      "end": 516.263,
      "text": "打开它的GitHub页面",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14765555171643274,
      "no_speech_prob": 9.698554459536268e-12,
      "compression_ratio": 1.072289156626506
    },
    {
      "start": 516.263,
      "end": 518.343,
      "text": "点击Releases",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14765555171643274,
      "no_speech_prob": 9.698554459536268e-12,
      "compression_ratio": 1.072289156626506
    },
    {
      "start": 518.343,
      "end": 521.263,
      "text": "下载CPU版本的压缩包",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14765555171643274,
      "no_speech_prob": 9.698554459536268e-12,
      "compression_ratio": 1.072289156626506
    },
    {
      "start": 521.263,
      "end": 524.103,
      "text": "解压后里面没有音频模型",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14765555171643274,
      "no_speech_prob": 9.698554459536268e-12,
      "compression_ratio": 1.072289156626506
    },
    {
      "start": 524.103,
      "end": 528.463,
      "text": "我们需要自己创建一个Models玩件夹",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14765555171643274,
      "no_speech_prob": 9.698554459536268e-12,
      "compression_ratio": 1.072289156626506
    },
    {
      "start": 528.463,
      "end": 530.863,
      "text": "用来存放所有模型文件",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14765555171643274,
      "no_speech_prob": 9.698554459536268e-12,
      "compression_ratio": 1.072289156626506
    },
    {
      "start": 530.863,
      "end": 534.803,
      "text": "先安装Pocket TTS",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14765555171643274,
      "no_speech_prob": 9.698554459536268e-12,
      "compression_ratio": 1.072289156626506
    },
    {
      "start": 534.803,
      "end": 537.943,
      "text": "打开Hugging Face上的模型页面",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14765555171643274,
      "no_speech_prob": 9.698554459536268e-12,
      "compression_ratio": 1.072289156626506
    },
    {
      "start": 537.943,
      "end": 539.583,
      "text": "它只有1亿参数",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15255254857680378,
      "no_speech_prob": 1.0617524220934982e-11,
      "compression_ratio": 1.0234741784037558
    },
    {
      "start": 539.583,
      "end": 540.383,
      "text": "非常轻量",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15255254857680378,
      "no_speech_prob": 1.0617524220934982e-11,
      "compression_ratio": 1.0234741784037558
    },
    {
      "start": 540.383,
      "end": 542.983,
      "text": "不用GPU也能流畅运行",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15255254857680378,
      "no_speech_prob": 1.0617524220934982e-11,
      "compression_ratio": 1.0234741784037558
    },
    {
      "start": 542.983,
      "end": 545.363,
      "text": "下载英文相关的模型文件",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15255254857680378,
      "no_speech_prob": 1.0617524220934982e-11,
      "compression_ratio": 1.0234741784037558
    },
    {
      "start": 545.363,
      "end": 547.323,
      "text": "分磁器和参考音色",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15255254857680378,
      "no_speech_prob": 1.0617524220934982e-11,
      "compression_ratio": 1.0234741784037558
    },
    {
      "start": 547.323,
      "end": 554.423,
      "text": "在Models下",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15255254857680378,
      "no_speech_prob": 1.0617524220934982e-11,
      "compression_ratio": 1.0234741784037558
    },
    {
      "start": 554.423,
      "end": 556.583,
      "text": "创建Pocket TS文件夹",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15255254857680378,
      "no_speech_prob": 1.0617524220934982e-11,
      "compression_ratio": 1.0234741784037558
    },
    {
      "start": 556.583,
      "end": 557.863,
      "text": "把文件放进去",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15255254857680378,
      "no_speech_prob": 1.0617524220934982e-11,
      "compression_ratio": 1.0234741784037558
    },
    {
      "start": 557.863,
      "end": 562.843,
      "text": "再安装QVN3 TTS",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15255254857680378,
      "no_speech_prob": 1.0617524220934982e-11,
      "compression_ratio": 1.0234741784037558
    },
    {
      "start": 562.843,
      "end": 564.783,
      "text": "打开摩塔社区",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15255254857680378,
      "no_speech_prob": 1.0617524220934982e-11,
      "compression_ratio": 1.0234741784037558
    },
    {
      "start": 564.783,
      "end": 566.703,
      "text": "搜索QVN3 TTS",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15255254857680378,
      "no_speech_prob": 1.0617524220934982e-11,
      "compression_ratio": 1.0234741784037558
    },
    {
      "start": 566.703,
      "end": 573.963,
      "text": "点击下载模型",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15824242642051295,
      "no_speech_prob": 7.829571192774498e-12,
      "compression_ratio": 1.0520231213872833
    },
    {
      "start": 573.963,
      "end": 577.983,
      "text": "推荐用GitLFS工具来下载",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15824242642051295,
      "no_speech_prob": 7.829571192774498e-12,
      "compression_ratio": 1.0520231213872833
    },
    {
      "start": 577.983,
      "end": 580.623,
      "text": "它是管理大文件的扩展",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15824242642051295,
      "no_speech_prob": 7.829571192774498e-12,
      "compression_ratio": 1.0520231213872833
    },
    {
      "start": 580.623,
      "end": 583.123,
      "text": "复制GitClone地址",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15824242642051295,
      "no_speech_prob": 7.829571192774498e-12,
      "compression_ratio": 1.0520231213872833
    },
    {
      "start": 583.123,
      "end": 586.043,
      "text": "在模型文件夹里打开CMD",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15824242642051295,
      "no_speech_prob": 7.829571192774498e-12,
      "compression_ratio": 1.0520231213872833
    },
    {
      "start": 586.043,
      "end": 587.283,
      "text": "粘贴并回车",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15824242642051295,
      "no_speech_prob": 7.829571192774498e-12,
      "compression_ratio": 1.0520231213872833
    },
    {
      "start": 587.283,
      "end": 591.483,
      "text": "如果已经下载过",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15824242642051295,
      "no_speech_prob": 7.829571192774498e-12,
      "compression_ratio": 1.0520231213872833
    },
    {
      "start": 591.483,
      "end": 592.563,
      "text": "会提示已存在",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15824242642051295,
      "no_speech_prob": 7.829571192774498e-12,
      "compression_ratio": 1.0520231213872833
    },
    {
      "start": 592.563,
      "end": 597.503,
      "text": "你也可以从Hugging Face的科问仓库下载",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11108014077851266,
      "no_speech_prob": 8.433902014270966e-12,
      "compression_ratio": 0.9908675799086758
    },
    {
      "start": 597.503,
      "end": 603.223,
      "text": "对于不想折腾的小伙伴",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11108014077851266,
      "no_speech_prob": 8.433902014270966e-12,
      "compression_ratio": 0.9908675799086758
    },
    {
      "start": 603.223,
      "end": 604.743,
      "text": "官方也提供了整合包",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11108014077851266,
      "no_speech_prob": 8.433902014270966e-12,
      "compression_ratio": 0.9908675799086758
    },
    {
      "start": 604.743,
      "end": 609.543,
      "text": "包含CPU、GPU和Web UI版本",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11108014077851266,
      "no_speech_prob": 8.433902014270966e-12,
      "compression_ratio": 0.9908675799086758
    },
    {
      "start": 609.543,
      "end": 611.843,
      "text": "常用音频模型都集成好了",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11108014077851266,
      "no_speech_prob": 8.433902014270966e-12,
      "compression_ratio": 0.9908675799086758
    },
    {
      "start": 611.843,
      "end": 613.783,
      "text": "直接用CPU就能跑",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11108014077851266,
      "no_speech_prob": 8.433902014270966e-12,
      "compression_ratio": 0.9908675799086758
    },
    {
      "start": 613.783,
      "end": 618.923,
      "text": "好,咱们快速回顾一下",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11108014077851266,
      "no_speech_prob": 8.433902014270966e-12,
      "compression_ratio": 0.9908675799086758
    },
    {
      "start": 618.923,
      "end": 623.263,
      "text": "Audio CPP让本地语音AI部署变得非常简单",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.08447621873587616,
      "no_speech_prob": 1.0493307611714187e-11,
      "compression_ratio": 1.051383399209486
    },
    {
      "start": 623.263,
      "end": 631.143,
      "text": "CLI和Web UI都支持涵盖TTS、特隆、ASR转写音乐生成和实时对话",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.08447621873587616,
      "no_speech_prob": 1.0493307611714187e-11,
      "compression_ratio": 1.051383399209486
    },
    {
      "start": 631.143,
      "end": 633.763,
      "text": "提供了统一的语音服务接口",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.08447621873587616,
      "no_speech_prob": 1.0493307611714187e-11,
      "compression_ratio": 1.051383399209486
    },
    {
      "start": 633.763,
      "end": 636.483,
      "text": "安装也只需要下载模型文件即可",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.08447621873587616,
      "no_speech_prob": 1.0493307611714187e-11,
      "compression_ratio": 1.051383399209486
    },
    {
      "start": 636.483,
      "end": 638.263,
      "text": "感兴趣的话快去试试吧",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.08447621873587616,
      "no_speech_prob": 1.0493307611714187e-11,
      "compression_ratio": 1.051383399209486
    },
    {
      "start": 638.263,
      "end": 639.063,
      "text": "感谢观看",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.08447621873587616,
      "no_speech_prob": 1.0493307611714187e-11,
      "compression_ratio": 1.051383399209486
    },
    {
      "start": 639.063,
      "end": 640.203,
      "text": "我们下期再见",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.08447621873587616,
      "no_speech_prob": 1.0493307611714187e-11,
      "compression_ratio": 1.051383399209486
    }
  ],
  "segmented_transcription": {
    "status": "SEGMENTED_TRANSCRIPT_OK",
    "created_at": "2026-08-01T17:06:25",
    "manifest": "/Users/sagawa/AI_WORK/video_notes/url/20260801_170556__audio.cpp_音频_AI_领域的_Ollama_本地部署音频AI模型_开箱即用_支持命令行_CLI_以及Web_UI_TTS_声音克隆_ASR/segmented_work/chunk_manifest.json",
    "audio": "/Users/sagawa/AI_WORK/video_notes/url/20260801_170556__audio.cpp_音频_AI_领域的_Ollama_本地部署音频AI模型_开箱即用_支持命令行_CLI_以及Web_UI_TTS_声音克隆_ASR/audio_16k_mono.wav",
    "audio_sha256": "332453776acf3e41c9931436710a6d5bbc14d3aadf6e2535b24c41ffa4aabf6d",
    "audio_duration": 645.445063,
    "model": "mlx-community/whisper-large-v3-turbo",
    "language_hint": "zh",
    "condition_on_previous_text": false,
    "chunk_count": 2,
    "resumed_chunks": 0,
    "merged_segment_count": 214,
    "final_segment_end": 640.203,
    "tail_repeat_run": 1,
    "tail_repeat_text": "下载英文相关的模型文件",
    "qc_ok": true,
    "qc_errors": []
  }
}