{
  "text": "大家好这里是 Hex-电脑课堂\n本期要跟大家分享的是\n一款开源的同声传译的神器\n那这款工具在我自己使用体验后\n我发现它的效果真的很惊艳\n它不只是跨平台\n支持Windows Linux还有macos\n并且整个过程是完全免费的\n更重要的是\n它的效果是\n我觉得是可以跟很多\n商用闭源的去扳扳手腕的\nOK那我们废话不多说\n先给大家直接展示效果\n人工智能\n是一项对社会影响非常深远的技术\n它将重塑社会的各行各业\nOK那我们从现在开始\n就要密切的关注这项技术并且\n保持与时俱进\n所以关注HEX电脑课堂\n并且持续的学习\n让我们一起在AI时代保持进步\n人工智能是一项对社会\n发展影响深远的技术\n它将重塑我们社会各行各业\n人工智能\n它是一项对社会发展影响深远的技术\n它将重塑我们社会各行各业\nOK那大家可以看到刚才的演示\n是不是它的效果非常好\n那本期\n我就将带大家手把手的来\n完成这样项目的一个部署OK\n那首先\n它对系统是有些要求的\n就是你是win10，win11\n这个都没关系\n然后你的电脑如果\n没有GPU的话\n也是ok的\n它要求比较低\n但是如果你有GPU\n有英伟达的显卡\n那更好了\n那建议显存是大于等于8G以上\n然后你的内存最低要4G\n这是最低的\n然后你的磁盘空间的话\n是200兆以上\n这个都是很低的条件\n然后你的麦克风和\n扬声器要正常工作\n第一步我们去下载这个终端\n我们来到这个网页\n这他的官方网页\n我们点击这个下载\n然后点击下载之后\n这个终端就会来\n下载到有这样一个exe的文件\n然后我们双击它\n然后点击下一步下一步安装\n就可以安装到本地了OK\n然后我顺便提一下\n因为它有两种形式\n我们今天讲的是这种终端的形式\n还有一种是它浏览器的\n一个拓展插件的形式\n那这个后面有机会我再给大家讲\n那ok你终端装好之后\n你桌面就会有这样一个LOGO图标\n那我们现在双击打开它\n它就会来到这样一个页面\n我们可以看到右上角有\n登录有字幕有设置等按钮\n那我们这里\n点击这个设置\n我们这里可以选择我说什么语言\n我这里选中文\n然后对方听到什么语言\n你可以去选的\n就是这里选英语\n那你一开始进来这里\n都是没有什么东西的\n你要自己去选\n然后这些选的话也\n不是你可以任意去选的\n要根据你自己去配置的模型\n那模型具体怎么配置\n我跟大家说一下\n比如大家可以看到\n它这里总共是包括了三个模块\n这里比如说有语音识别\n有语音的那个文本的翻译\n还有语音的合成\n那这几个模型是我自己下载的\n那我为什么选择下载这几个模型\n我给大家说一下这个原因\n因为我自己做了很多的一些测试OK\n那这第二步的话\n我们本期先不讲\n因为这个东西是跟插件相关的\n跟我们这期内容不是很相关\n所以我们这边先过\n然后我们先看到第三步\n就是模型下载这里\n模型下载这里了\n首先它分成三个模型\n刚刚跟大家提过了\n然后我建议大家去下载哪个模型\n去下载这一个\n这个模型\n这个mini 4B的\n你可以看到为什么\n因为这个模型\n它的整体的效果是最好的\n在这个地方你可以看到\n它不管是它的精度还是速度\n它都是5颗星\n是满星的\n所以我这边是给\n用这个模型给大家去演示\n然后我等下会教大家怎么去下载\n对然后\n我顺便给大家提一下\n它这里有提到一个概念\n就是什么webgpu\n这个东西就是相当于是用GPU去做\n网页端的一个渲染加速的\n以往都是CPU\n所以你选它的话会更好\n速度更快\n第二步的话去下载机器翻译的模型\n那机器翻译的模型官方\n这边推荐我们是用这个模型\n就是这个叫做translate Gemma\n它的速度和质量整体也是最好的\n所以我们就选它\n然后这边下载的好了之后\n这个模型大概是有将近3G左右\n然后你可以看到它这边\n支持了各种各样的语言\n语种也比较丰富\nOK\n然后TTS模型就是语音合成\n就是那个播音的\n那这个的话\n我这边是选的是这个supersonic\n那这个模型我们可以\n看到它只有380多兆\n对吧它比较小\n然后它支持的语种也很丰富\n那这样的话\n就是你组合起来就会非常的丰富了\n对吧你语音识别跟语音合成\n包括那个翻译\n那语种的范围就很丰富了\n你的可拓展性就很强了对吧\nOK那它怎么样配来\n教大家首先我们来到语音识别这里\n你默认的话它是会\n这里有一个叫做浏览模型库\n我们点击一下\n然后这边有很多的模型\n大家可以看到\n这边我们就选择刚才已经下好的\n我这边已经下好了\n它就会显示使用中\n你没有下载好的\n我举个例子\n你就点一下这个下载按钮\n你点一下下载\n它就会给你开始在本地去下载了\n然后你就只要等个几分钟就好了\n那这里的话\n需要注意的就是你的\n网络一定要保持全局模式\n网络要稳定\n你看他就现在已经开始下载了OK\n那我这边暂时不用给他下\n我就给他叉掉\n我还是依然用我们的\n这个mini的4B的这个模型OK然后\n我们再接着去找刚才另外几个模型\n比如说\n语音那个翻译的模型\n这个也有\n这里我们\n点击这个浏览模型库\n都是一样的\n很容易操作\n这边我已经下载好了这个translate Gemma\n刚才那个4B模型\n那其他的模型你要下载也是一样的\n在这个地方点下载按钮就好了\n很简单\n然后最后一个这个模型就是\n那个语音合成的这个模型\n同样还是在这里点浏览模型库\n可以看到这底下有很多待选的\n你根据自己的需求来选\n那它这边会有一些标签\n就是什么en, fa这些\n这些是代表你这个语音的模型\n它支持哪些语种\n我这里选择一个很多的\n大家可以看到\n我把它也下载好了\n下载好之后\n这三个模型都配好了OK\n那我们现在要把它选中了\n就是你已经把它都选好了\n然后这里有语音合成的一个速度\n这个东西\n它默认是1.0\n我这里给它改成了1.2\n可能我觉得稍微速度会快一点\n这个根据个人的一个需求吧\n都没有一个绝对的标准\n然后底下这几个参数\n我要给大家稍微提一下\n这些都是默认已经配好的\n它都是一个比较成熟的参数\n你不用一般不用去改的\n像这个什么\n比如说这个最小的静音时长\n这一项它一般是设置为1.4秒\n是什么意思\n就是我们说话的时候\n如果两句话间隔超过了1.4秒的话\n那么它就会把这里切割开\n分成两句话\n就这意思\n然后最小的语音时长\n它这里默认写的是0.4秒\n也就是说\n如果你说话的有一些\n那种声音比较零碎\n那个声音的片段小于0.4秒\n他就不会给你去识别\n就相当于是杂音对吧\n这也是比较节省算力的一种方法\nOK那这个都弄好之后\n我们这里还有一个音频\n这一项也是你的\n就是你的物理的外设\n我们要选什么\n首先你的麦克风\n你要选你的实际的物理的外设\n你不要选其他的\n不然他到时候连不通的\n然后\n这个东西其他一般就不用管了\n然后还有这个供提供商\n这个刚刚我们已经说过了\n对然后就是常规这里\n然后你要说什么\n你要说中文\n然后你希望对方听到什么\n我选英文\n那你这里的语言你要注意\n因为我刚才选的那些模型\n它支持了很多种语种\n所以我这里才有的这么多选\n如果你选的模型它就\n支持一两种语种的话\n你这个下拉框是没这么多选的\n那我们这边选好之后\n我们就点击这里开始会话\n然后左边\n它是有3个\n这里是我\n还有一个对方\n还有一个两者\n那我们目前主要是用\n我这种方式就好了\n如果是对方的话\n那就是另外一种\n这后面这两个我们今天不演示了OK\n然后我们点击开始会话\n这里加载好了之后\n我们现在就可以说话了\n然后你可以看到左上角\n它正在识别我们讲的话\n然后它这个是实时的\n是流式的\n那这个字有点小\n我们可以点一下右上角\n这里有一个字幕\n把它放大一些\n对整体看上去\n它这个效果还可以\n虽然说有时候可能\n识别的个别字不太精准\n那当然这可能也是我个人\n有时候普通话不太标准\n所以它识别起来\n可能会有一些点误差\n但是也不影响它的一个翻译\n我们可以看到它整体\n我觉得效果还是不错的\nOK我们退出的话\n我们就点一下ESC\n然后这些字幕的话\n它有一个删除的小按钮点一下\n它就给它清空了\n然后我们要结束的话\n我们点击下方的这个结束会话\n那这个就结束了\nOK那如果你希望对方听到的\n就我刚刚给大家演示了不同的语种\n这个地方比如说你\n想要他听到西班牙语\n或者是你想要他听到日本语\n那你这个地方切换一下就好了\n然后你这里点击开始会话\n你好可以听到我声音吗\n你好\n我想现在跟你分享一些事情\n你可以看到\n他这个字幕还是\n就是逐步的播过去的\n这个我觉得效果是非常好的\n就是有你讲话\n然后你可以看到这个字幕是在流动的\nOK然后我们把它结束\n然后我们可以再切换其他的这种语言\n那这些都是类似的\n我就不给大家一一去演示了\n然后这上面有一些东西\n我再给大家讲一下\n就是提供商\n那因为本地的其实我已经试过了\n这几个模型它都是在本地\n就没有去走任何在线的API\n那如果有些小伙伴他说\n我有在线的这种API的话\n你也可以去用\n这个都行\n因为我觉得其实本地\n的效果已经很棒了\n大家可以看到\n那这边怎么选\n你可以看到\n首先在这个地方点常规\n然后这里有个AI服务提供商\n我们点一下\n这里有Openai\n有一些其他的\n比如说什么豆包\n还有那个谷歌Gemini呀\n等等的这些\n你有买哪个API\n你就给它对应的输入\n比如我点Openai的\n你看你点进Openai的\n这个地方填入你的密钥\n然后这里音频的话\n这个保持不动\n提供商这一块的话\n你可以选\n在这里选择\n就是Openai那个语音的模型对应的\n发音的那个角色对吧\n你自己去选\n然后这些东西你可以把\n它参数对应的填过去\n它就可以调在线的API了，对\n那原理是一样的\n但是我自己是觉得\n就是其实我们免费的\n就是我今天给大家\n展示的主要是免费的\n已经够用了对\n所以整体实现下来这个效果是\n我觉得是在我目前\n就是看过的所有的开源的\n这种同声传译的项目里面\n应该是最好的了\n所以我强烈建议大家\n尤其是很多小伙伴经常需要\n就是用到外语去交流\n但是自己可能外语\n水平可能还没那么好\n用这些工具我觉得是\n对你有很大的帮助的OK\n那本期的这些课程资源其实比较少\n但是我会放到评论区\n大家直接可以去用\n那我最后还要提到一个注意点\n就是我在实际使用过程中\n很多小伙伴使用的时候会报错\n那这个报错是什么原因造成的\n就是我们一定要注意\n我们用到的这三个模型\n它支持的语种范围是什么\n举个例子\n比如说我这里的语种\n它有一个粤语\n这也是支持的\n我们可以看到这有粤语对吧但是\n我这里说话的话\n我说的是普通话然后\n我这个语音识别可能它这个模型\n它能够识别普通话\n但是它语音合成的那个模型\n它不支持播放粤语\n那么它就会报错\n你这三个模型\n大家可以看到\n就是你这三个模型\n它们支持的语种一定要有共同的\n那么这个链路才是通的\n有一个不支持的\n那么你这个链路就通不了\n小伙伴可能遇到的这些\n大部分可能是这一方面的问题\n还有就是你网络不好\n你网络如果有问题的话\n你没开全局模式的话\n你这些模型是下载不到你本地的\n还有一个细节就是\n关于这里有个登录\n这里刚才跟大家有讲过\n你看这个登录这里\n你可以点击去注册\n或者是就已经有账号就登录进去\n那它本质上\n它会用到它云端的这个sokuji\n它自己的翻译服务 api\n那对于有些小伙伴不想去折腾\n你可以去用它\n那这样的话你就不用\n去本地下载模型了\n但是要提醒一下大家就是\n你这样的话\n你的数据是会经过它的服务器的\n就是说数据的隐私性\n就保障没那么好\n根据自己的需求来\n如果你全都想让数据不\n出自己的本地的电脑\n那么你就用我刚刚给大家讲的方法\n还有一点\n也是我个人使用的一个经验\n这个东西对大家应该很有帮助\n同声传译这个东西\n大家要知道它最看重的是什么\n就是一个是翻译的准确性\n一个是它的一个时延\n就是我们那个翻译的\n那个时间不能太久了\n这样的话就做不到同声了\n那我们要想在这个项目中\n要尽量的去把它时延降低\n让它的用户体验更好\n应该怎么去做\n首先刚才我跟大家是讲了\n它里面有很多模型对吧\n大家可以看到这模型是有一大堆的\n那么我们应该怎么去选模型\n大家要记住\n你要选肯定要去选流式的\n也就是说不要选非流式的模型\n那流式的模型一般会带上\n一个叫realtime的这个字段\n然后有这个Web GPU的\n你就尽量去选它对\n这个是一个细节\n很重要的\n就是你选那种非流式的\n它推你的话\n那整个链路就会很慢了\n等于说你说完半句话以后\n别人都已经说完下一句了\n它还没翻译完\n那当然那那体验就很不好了对吧\n所以大家一定要记住\n就是要尽量选流式的模型\n不管是语音识别\n还是那个翻译\n或者是那个语音合成\n都要去选流式的\n然后有GPU的话\n就要去把那个webgpu那个也选好\n那本期视频\n我就跟大家分享到这了\n大家如果觉得我的\n视频对你有帮助的话\n也请支持一下我的创作\n帮我点赞关注转发\n那本期内容到此结束\n我们下期再会",
  "language": "zh",
  "duration": 930.795125,
  "segments": [
    {
      "start": 0.533,
      "end": 2.6,
      "text": "大家好这里是 Hex-电脑课堂",
      "source": "source_caption"
    },
    {
      "start": 2.666,
      "end": 4.0,
      "text": "本期要跟大家分享的是",
      "source": "source_caption"
    },
    {
      "start": 4.0,
      "end": 6.566,
      "text": "一款开源的同声传译的神器",
      "source": "source_caption"
    },
    {
      "start": 6.6,
      "end": 9.1,
      "text": "那这款工具在我自己使用体验后",
      "source": "source_caption"
    },
    {
      "start": 9.1,
      "end": 10.8,
      "text": "我发现它的效果真的很惊艳",
      "source": "source_caption"
    },
    {
      "start": 11.066,
      "end": 13.066,
      "text": "它不只是跨平台",
      "source": "source_caption"
    },
    {
      "start": 13.066,
      "end": 15.533,
      "text": "支持Windows Linux还有macos",
      "source": "source_caption"
    },
    {
      "start": 15.666,
      "end": 17.7,
      "text": "并且整个过程是完全免费的",
      "source": "source_caption"
    },
    {
      "start": 17.866,
      "end": 18.466,
      "text": "更重要的是",
      "source": "source_caption"
    },
    {
      "start": 18.466,
      "end": 19.3,
      "text": "它的效果是",
      "source": "source_caption"
    },
    {
      "start": 19.3,
      "end": 20.6,
      "text": "我觉得是可以跟很多",
      "source": "source_caption"
    },
    {
      "start": 20.6,
      "end": 22.666,
      "text": "商用闭源的去扳扳手腕的",
      "source": "source_caption"
    },
    {
      "start": 22.666,
      "end": 23.933,
      "text": "OK那我们废话不多说",
      "source": "source_caption"
    },
    {
      "start": 23.933,
      "end": 25.4,
      "text": "先给大家直接展示效果",
      "source": "source_caption"
    },
    {
      "start": 26.466,
      "end": 27.4,
      "text": "人工智能",
      "source": "source_caption"
    },
    {
      "start": 27.4,
      "end": 30.166,
      "text": "是一项对社会影响非常深远的技术",
      "source": "source_caption"
    },
    {
      "start": 30.266,
      "end": 32.4,
      "text": "它将重塑社会的各行各业",
      "source": "source_caption"
    },
    {
      "start": 45.0,
      "end": 46.466,
      "text": "OK那我们从现在开始",
      "source": "source_caption"
    },
    {
      "start": 46.466,
      "end": 49.1,
      "text": "就要密切的关注这项技术并且",
      "source": "source_caption"
    },
    {
      "start": 49.1,
      "end": 51.3,
      "text": "保持与时俱进",
      "source": "source_caption"
    },
    {
      "start": 60.566,
      "end": 62.933,
      "text": "所以关注HEX电脑课堂",
      "source": "source_caption"
    },
    {
      "start": 63.066,
      "end": 64.3,
      "text": "并且持续的学习",
      "source": "source_caption"
    },
    {
      "start": 64.3,
      "end": 66.6,
      "text": "让我们一起在AI时代保持进步",
      "source": "source_caption"
    },
    {
      "start": 76.933,
      "end": 78.866,
      "text": "人工智能是一项对社会",
      "source": "source_caption"
    },
    {
      "start": 78.866,
      "end": 80.766,
      "text": "发展影响深远的技术",
      "source": "source_caption"
    },
    {
      "start": 80.9,
      "end": 83.566,
      "text": "它将重塑我们社会各行各业",
      "source": "source_caption"
    },
    {
      "start": 97.2,
      "end": 98.0,
      "text": "人工智能",
      "source": "source_caption"
    },
    {
      "start": 98.066,
      "end": 101.066,
      "text": "它是一项对社会发展影响深远的技术",
      "source": "source_caption"
    },
    {
      "start": 101.133,
      "end": 103.6,
      "text": "它将重塑我们社会各行各业",
      "source": "source_caption"
    },
    {
      "start": 117.6,
      "end": 119.866,
      "text": "OK那大家可以看到刚才的演示",
      "source": "source_caption"
    },
    {
      "start": 119.866,
      "end": 121.566,
      "text": "是不是它的效果非常好",
      "source": "source_caption"
    },
    {
      "start": 121.8,
      "end": 122.9,
      "text": "那本期",
      "source": "source_caption"
    },
    {
      "start": 122.9,
      "end": 124.866,
      "text": "我就将带大家手把手的来",
      "source": "source_caption"
    },
    {
      "start": 124.966,
      "end": 126.8,
      "text": "完成这样项目的一个部署OK",
      "source": "source_caption"
    },
    {
      "start": 126.9,
      "end": 127.6,
      "text": "那首先",
      "source": "source_caption"
    },
    {
      "start": 127.6,
      "end": 129.1,
      "text": "它对系统是有些要求的",
      "source": "source_caption"
    },
    {
      "start": 129.1,
      "end": 130.6,
      "text": "就是你是win10，win11",
      "source": "source_caption"
    },
    {
      "start": 130.6,
      "end": 131.6,
      "text": "这个都没关系",
      "source": "source_caption"
    },
    {
      "start": 131.6,
      "end": 133.0,
      "text": "然后你的电脑如果",
      "source": "source_caption"
    },
    {
      "start": 133.133,
      "end": 134.333,
      "text": "没有GPU的话",
      "source": "source_caption"
    },
    {
      "start": 134.333,
      "end": 135.366,
      "text": "也是ok的",
      "source": "source_caption"
    },
    {
      "start": 135.366,
      "end": 136.5,
      "text": "它要求比较低",
      "source": "source_caption"
    },
    {
      "start": 136.533,
      "end": 138.2,
      "text": "但是如果你有GPU",
      "source": "source_caption"
    },
    {
      "start": 138.2,
      "end": 139.166,
      "text": "有英伟达的显卡",
      "source": "source_caption"
    },
    {
      "start": 139.166,
      "end": 139.766,
      "text": "那更好了",
      "source": "source_caption"
    },
    {
      "start": 139.766,
      "end": 142.566,
      "text": "那建议显存是大于等于8G以上",
      "source": "source_caption"
    },
    {
      "start": 142.766,
      "end": 146.2,
      "text": "然后你的内存最低要4G",
      "source": "source_caption"
    },
    {
      "start": 146.2,
      "end": 146.933,
      "text": "这是最低的",
      "source": "source_caption"
    },
    {
      "start": 146.933,
      "end": 148.866,
      "text": "然后你的磁盘空间的话",
      "source": "source_caption"
    },
    {
      "start": 148.866,
      "end": 150.533,
      "text": "是200兆以上",
      "source": "source_caption"
    },
    {
      "start": 150.533,
      "end": 151.9,
      "text": "这个都是很低的条件",
      "source": "source_caption"
    },
    {
      "start": 151.9,
      "end": 153.066,
      "text": "然后你的麦克风和",
      "source": "source_caption"
    },
    {
      "start": 153.066,
      "end": 154.533,
      "text": "扬声器要正常工作",
      "source": "source_caption"
    },
    {
      "start": 154.566,
      "end": 156.1,
      "text": "第一步我们去下载这个终端",
      "source": "source_caption"
    },
    {
      "start": 156.1,
      "end": 157.333,
      "text": "我们来到这个网页",
      "source": "source_caption"
    },
    {
      "start": 157.333,
      "end": 158.533,
      "text": "这他的官方网页",
      "source": "source_caption"
    },
    {
      "start": 158.533,
      "end": 160.0,
      "text": "我们点击这个下载",
      "source": "source_caption"
    },
    {
      "start": 160.266,
      "end": 161.6,
      "text": "然后点击下载之后",
      "source": "source_caption"
    },
    {
      "start": 161.6,
      "end": 162.8,
      "text": "这个终端就会来",
      "source": "source_caption"
    },
    {
      "start": 163.1,
      "end": 165.566,
      "text": "下载到有这样一个exe的文件",
      "source": "source_caption"
    },
    {
      "start": 165.566,
      "end": 166.5,
      "text": "然后我们双击它",
      "source": "source_caption"
    },
    {
      "start": 166.5,
      "end": 168.266,
      "text": "然后点击下一步下一步安装",
      "source": "source_caption"
    },
    {
      "start": 168.266,
      "end": 170.5,
      "text": "就可以安装到本地了OK",
      "source": "source_caption"
    },
    {
      "start": 170.533,
      "end": 171.766,
      "text": "然后我顺便提一下",
      "source": "source_caption"
    },
    {
      "start": 172.0,
      "end": 173.266,
      "text": "因为它有两种形式",
      "source": "source_caption"
    },
    {
      "start": 173.266,
      "end": 175.5,
      "text": "我们今天讲的是这种终端的形式",
      "source": "source_caption"
    },
    {
      "start": 175.5,
      "end": 176.9,
      "text": "还有一种是它浏览器的",
      "source": "source_caption"
    },
    {
      "start": 176.9,
      "end": 178.4,
      "text": "一个拓展插件的形式",
      "source": "source_caption"
    },
    {
      "start": 178.4,
      "end": 180.533,
      "text": "那这个后面有机会我再给大家讲",
      "source": "source_caption"
    },
    {
      "start": 180.566,
      "end": 182.966,
      "text": "那ok你终端装好之后",
      "source": "source_caption"
    },
    {
      "start": 182.966,
      "end": 185.066,
      "text": "你桌面就会有这样一个LOGO图标",
      "source": "source_caption"
    },
    {
      "start": 185.066,
      "end": 186.6,
      "text": "那我们现在双击打开它",
      "source": "source_caption"
    },
    {
      "start": 186.733,
      "end": 188.066,
      "text": "它就会来到这样一个页面",
      "source": "source_caption"
    },
    {
      "start": 188.066,
      "end": 189.3,
      "text": "我们可以看到右上角有",
      "source": "source_caption"
    },
    {
      "start": 189.3,
      "end": 191.533,
      "text": "登录有字幕有设置等按钮",
      "source": "source_caption"
    },
    {
      "start": 191.566,
      "end": 192.533,
      "text": "那我们这里",
      "source": "source_caption"
    },
    {
      "start": 192.533,
      "end": 193.733,
      "text": "点击这个设置",
      "source": "source_caption"
    },
    {
      "start": 194.266,
      "end": 196.666,
      "text": "我们这里可以选择我说什么语言",
      "source": "source_caption"
    },
    {
      "start": 196.666,
      "end": 197.6,
      "text": "我这里选中文",
      "source": "source_caption"
    },
    {
      "start": 197.6,
      "end": 198.966,
      "text": "然后对方听到什么语言",
      "source": "source_caption"
    },
    {
      "start": 198.966,
      "end": 199.8,
      "text": "你可以去选的",
      "source": "source_caption"
    },
    {
      "start": 199.8,
      "end": 201.2,
      "text": "就是这里选英语",
      "source": "source_caption"
    },
    {
      "start": 201.366,
      "end": 203.066,
      "text": "那你一开始进来这里",
      "source": "source_caption"
    },
    {
      "start": 203.066,
      "end": 204.1,
      "text": "都是没有什么东西的",
      "source": "source_caption"
    },
    {
      "start": 204.1,
      "end": 205.0,
      "text": "你要自己去选",
      "source": "source_caption"
    },
    {
      "start": 205.0,
      "end": 206.166,
      "text": "然后这些选的话也",
      "source": "source_caption"
    },
    {
      "start": 206.166,
      "end": 207.6,
      "text": "不是你可以任意去选的",
      "source": "source_caption"
    },
    {
      "start": 207.6,
      "end": 209.466,
      "text": "要根据你自己去配置的模型",
      "source": "source_caption"
    },
    {
      "start": 209.9,
      "end": 210.966,
      "text": "那模型具体怎么配置",
      "source": "source_caption"
    },
    {
      "start": 210.966,
      "end": 212.133,
      "text": "我跟大家说一下",
      "source": "source_caption"
    },
    {
      "start": 212.2,
      "end": 213.8,
      "text": "比如大家可以看到",
      "source": "source_caption"
    },
    {
      "start": 213.933,
      "end": 216.366,
      "text": "它这里总共是包括了三个模块",
      "source": "source_caption"
    },
    {
      "start": 216.366,
      "end": 218.466,
      "text": "这里比如说有语音识别",
      "source": "source_caption"
    },
    {
      "start": 218.5,
      "end": 220.766,
      "text": "有语音的那个文本的翻译",
      "source": "source_caption"
    },
    {
      "start": 220.766,
      "end": 221.9,
      "text": "还有语音的合成",
      "source": "source_caption"
    },
    {
      "start": 222.266,
      "end": 224.366,
      "text": "那这几个模型是我自己下载的",
      "source": "source_caption"
    },
    {
      "start": 224.366,
      "end": 226.333,
      "text": "那我为什么选择下载这几个模型",
      "source": "source_caption"
    },
    {
      "start": 226.333,
      "end": 228.466,
      "text": "我给大家说一下这个原因",
      "source": "source_caption"
    },
    {
      "start": 228.466,
      "end": 230.933,
      "text": "因为我自己做了很多的一些测试OK",
      "source": "source_caption"
    },
    {
      "start": 231.1,
      "end": 232.3,
      "text": "那这第二步的话",
      "source": "source_caption"
    },
    {
      "start": 232.3,
      "end": 233.4,
      "text": "我们本期先不讲",
      "source": "source_caption"
    },
    {
      "start": 233.4,
      "end": 235.566,
      "text": "因为这个东西是跟插件相关的",
      "source": "source_caption"
    },
    {
      "start": 235.566,
      "end": 237.766,
      "text": "跟我们这期内容不是很相关",
      "source": "source_caption"
    },
    {
      "start": 237.766,
      "end": 239.0,
      "text": "所以我们这边先过",
      "source": "source_caption"
    },
    {
      "start": 239.4,
      "end": 241.0,
      "text": "然后我们先看到第三步",
      "source": "source_caption"
    },
    {
      "start": 241.0,
      "end": 242.4,
      "text": "就是模型下载这里",
      "source": "source_caption"
    },
    {
      "start": 242.533,
      "end": 243.7,
      "text": "模型下载这里了",
      "source": "source_caption"
    },
    {
      "start": 243.7,
      "end": 244.9,
      "text": "首先它分成三个模型",
      "source": "source_caption"
    },
    {
      "start": 244.9,
      "end": 246.133,
      "text": "刚刚跟大家提过了",
      "source": "source_caption"
    },
    {
      "start": 246.133,
      "end": 248.266,
      "text": "然后我建议大家去下载哪个模型",
      "source": "source_caption"
    },
    {
      "start": 248.266,
      "end": 249.333,
      "text": "去下载这一个",
      "source": "source_caption"
    },
    {
      "start": 249.333,
      "end": 250.133,
      "text": "这个模型",
      "source": "source_caption"
    },
    {
      "start": 250.333,
      "end": 252.6,
      "text": "这个mini 4B的",
      "source": "source_caption"
    },
    {
      "start": 252.6,
      "end": 253.666,
      "text": "你可以看到为什么",
      "source": "source_caption"
    },
    {
      "start": 253.7,
      "end": 254.5,
      "text": "因为这个模型",
      "source": "source_caption"
    },
    {
      "start": 254.733,
      "end": 257.366,
      "text": "它的整体的效果是最好的",
      "source": "source_caption"
    },
    {
      "start": 257.366,
      "end": 258.766,
      "text": "在这个地方你可以看到",
      "source": "source_caption"
    },
    {
      "start": 258.766,
      "end": 261.133,
      "text": "它不管是它的精度还是速度",
      "source": "source_caption"
    },
    {
      "start": 261.133,
      "end": 262.1,
      "text": "它都是5颗星",
      "source": "source_caption"
    },
    {
      "start": 262.1,
      "end": 263.3,
      "text": "是满星的",
      "source": "source_caption"
    },
    {
      "start": 263.366,
      "end": 264.866,
      "text": "所以我这边是给",
      "source": "source_caption"
    },
    {
      "start": 264.866,
      "end": 266.366,
      "text": "用这个模型给大家去演示",
      "source": "source_caption"
    },
    {
      "start": 266.5,
      "end": 268.4,
      "text": "然后我等下会教大家怎么去下载",
      "source": "source_caption"
    },
    {
      "start": 268.466,
      "end": 269.566,
      "text": "对然后",
      "source": "source_caption"
    },
    {
      "start": 269.566,
      "end": 270.966,
      "text": "我顺便给大家提一下",
      "source": "source_caption"
    },
    {
      "start": 270.966,
      "end": 272.2,
      "text": "它这里有提到一个概念",
      "source": "source_caption"
    },
    {
      "start": 272.2,
      "end": 274.4,
      "text": "就是什么webgpu",
      "source": "source_caption"
    },
    {
      "start": 274.6,
      "end": 276.966,
      "text": "这个东西就是相当于是用GPU去做",
      "source": "source_caption"
    },
    {
      "start": 276.966,
      "end": 278.9,
      "text": "网页端的一个渲染加速的",
      "source": "source_caption"
    },
    {
      "start": 278.9,
      "end": 279.8,
      "text": "以往都是CPU",
      "source": "source_caption"
    },
    {
      "start": 279.8,
      "end": 281.533,
      "text": "所以你选它的话会更好",
      "source": "source_caption"
    },
    {
      "start": 281.533,
      "end": 282.2,
      "text": "速度更快",
      "source": "source_caption"
    },
    {
      "start": 282.2,
      "end": 284.333,
      "text": "第二步的话去下载机器翻译的模型",
      "source": "source_caption"
    },
    {
      "start": 284.766,
      "end": 286.2,
      "text": "那机器翻译的模型官方",
      "source": "source_caption"
    },
    {
      "start": 286.2,
      "end": 287.8,
      "text": "这边推荐我们是用这个模型",
      "source": "source_caption"
    },
    {
      "start": 288.0,
      "end": 290.4,
      "text": "就是这个叫做translate Gemma",
      "source": "source_caption"
    },
    {
      "start": 290.6,
      "end": 292.766,
      "text": "它的速度和质量整体也是最好的",
      "source": "source_caption"
    },
    {
      "start": 292.766,
      "end": 294.3,
      "text": "所以我们就选它",
      "source": "source_caption"
    },
    {
      "start": 294.333,
      "end": 295.866,
      "text": "然后这边下载的好了之后",
      "source": "source_caption"
    },
    {
      "start": 295.866,
      "end": 298.266,
      "text": "这个模型大概是有将近3G左右",
      "source": "source_caption"
    },
    {
      "start": 298.333,
      "end": 299.333,
      "text": "然后你可以看到它这边",
      "source": "source_caption"
    },
    {
      "start": 299.333,
      "end": 300.9,
      "text": "支持了各种各样的语言",
      "source": "source_caption"
    },
    {
      "start": 300.9,
      "end": 301.866,
      "text": "语种也比较丰富",
      "source": "source_caption"
    },
    {
      "start": 301.866,
      "end": 302.666,
      "text": "OK",
      "source": "source_caption"
    },
    {
      "start": 302.766,
      "end": 305.133,
      "text": "然后TTS模型就是语音合成",
      "source": "source_caption"
    },
    {
      "start": 305.133,
      "end": 306.266,
      "text": "就是那个播音的",
      "source": "source_caption"
    },
    {
      "start": 306.3,
      "end": 307.133,
      "text": "那这个的话",
      "source": "source_caption"
    },
    {
      "start": 307.133,
      "end": 309.933,
      "text": "我这边是选的是这个supersonic",
      "source": "source_caption"
    },
    {
      "start": 310.0,
      "end": 311.066,
      "text": "那这个模型我们可以",
      "source": "source_caption"
    },
    {
      "start": 311.066,
      "end": 313.066,
      "text": "看到它只有380多兆",
      "source": "source_caption"
    },
    {
      "start": 313.5,
      "end": 314.666,
      "text": "对吧它比较小",
      "source": "source_caption"
    },
    {
      "start": 314.666,
      "end": 316.6,
      "text": "然后它支持的语种也很丰富",
      "source": "source_caption"
    },
    {
      "start": 316.733,
      "end": 317.7,
      "text": "那这样的话",
      "source": "source_caption"
    },
    {
      "start": 317.7,
      "end": 321.1,
      "text": "就是你组合起来就会非常的丰富了",
      "source": "source_caption"
    },
    {
      "start": 321.1,
      "end": 323.166,
      "text": "对吧你语音识别跟语音合成",
      "source": "source_caption"
    },
    {
      "start": 323.166,
      "end": 324.166,
      "text": "包括那个翻译",
      "source": "source_caption"
    },
    {
      "start": 324.166,
      "end": 326.1,
      "text": "那语种的范围就很丰富了",
      "source": "source_caption"
    },
    {
      "start": 326.1,
      "end": 328.333,
      "text": "你的可拓展性就很强了对吧",
      "source": "source_caption"
    },
    {
      "start": 328.366,
      "end": 330.2,
      "text": "OK那它怎么样配来",
      "source": "source_caption"
    },
    {
      "start": 330.466,
      "end": 332.733,
      "text": "教大家首先我们来到语音识别这里",
      "source": "source_caption"
    },
    {
      "start": 332.733,
      "end": 334.266,
      "text": "你默认的话它是会",
      "source": "source_caption"
    },
    {
      "start": 334.266,
      "end": 337.1,
      "text": "这里有一个叫做浏览模型库",
      "source": "source_caption"
    },
    {
      "start": 337.1,
      "end": 338.166,
      "text": "我们点击一下",
      "source": "source_caption"
    },
    {
      "start": 338.266,
      "end": 339.666,
      "text": "然后这边有很多的模型",
      "source": "source_caption"
    },
    {
      "start": 339.666,
      "end": 340.7,
      "text": "大家可以看到",
      "source": "source_caption"
    },
    {
      "start": 340.7,
      "end": 343.133,
      "text": "这边我们就选择刚才已经下好的",
      "source": "source_caption"
    },
    {
      "start": 343.133,
      "end": 343.933,
      "text": "我这边已经下好了",
      "source": "source_caption"
    },
    {
      "start": 343.933,
      "end": 345.6,
      "text": "它就会显示使用中",
      "source": "source_caption"
    },
    {
      "start": 345.6,
      "end": 346.8,
      "text": "你没有下载好的",
      "source": "source_caption"
    },
    {
      "start": 346.8,
      "end": 347.666,
      "text": "我举个例子",
      "source": "source_caption"
    },
    {
      "start": 347.666,
      "end": 349.1,
      "text": "你就点一下这个下载按钮",
      "source": "source_caption"
    },
    {
      "start": 349.3,
      "end": 350.266,
      "text": "你点一下下载",
      "source": "source_caption"
    },
    {
      "start": 350.266,
      "end": 352.266,
      "text": "它就会给你开始在本地去下载了",
      "source": "source_caption"
    },
    {
      "start": 352.4,
      "end": 354.5,
      "text": "然后你就只要等个几分钟就好了",
      "source": "source_caption"
    },
    {
      "start": 354.5,
      "end": 355.466,
      "text": "那这里的话",
      "source": "source_caption"
    },
    {
      "start": 355.466,
      "end": 356.7,
      "text": "需要注意的就是你的",
      "source": "source_caption"
    },
    {
      "start": 356.7,
      "end": 358.8,
      "text": "网络一定要保持全局模式",
      "source": "source_caption"
    },
    {
      "start": 358.8,
      "end": 359.6,
      "text": "网络要稳定",
      "source": "source_caption"
    },
    {
      "start": 359.6,
      "end": 361.733,
      "text": "你看他就现在已经开始下载了OK",
      "source": "source_caption"
    },
    {
      "start": 361.766,
      "end": 362.966,
      "text": "那我这边暂时不用给他下",
      "source": "source_caption"
    },
    {
      "start": 362.966,
      "end": 363.8,
      "text": "我就给他叉掉",
      "source": "source_caption"
    },
    {
      "start": 363.8,
      "end": 364.966,
      "text": "我还是依然用我们的",
      "source": "source_caption"
    },
    {
      "start": 364.966,
      "end": 368.5,
      "text": "这个mini的4B的这个模型OK然后",
      "source": "source_caption"
    },
    {
      "start": 368.666,
      "end": 371.366,
      "text": "我们再接着去找刚才另外几个模型",
      "source": "source_caption"
    },
    {
      "start": 371.5,
      "end": 372.3,
      "text": "比如说",
      "source": "source_caption"
    },
    {
      "start": 372.533,
      "end": 373.933,
      "text": "语音那个翻译的模型",
      "source": "source_caption"
    },
    {
      "start": 373.933,
      "end": 374.666,
      "text": "这个也有",
      "source": "source_caption"
    },
    {
      "start": 374.666,
      "end": 375.5,
      "text": "这里我们",
      "source": "source_caption"
    },
    {
      "start": 375.5,
      "end": 376.966,
      "text": "点击这个浏览模型库",
      "source": "source_caption"
    },
    {
      "start": 376.966,
      "end": 377.666,
      "text": "都是一样的",
      "source": "source_caption"
    },
    {
      "start": 377.666,
      "end": 378.8,
      "text": "很容易操作",
      "source": "source_caption"
    },
    {
      "start": 379.066,
      "end": 381.866,
      "text": "这边我已经下载好了这个translate Gemma",
      "source": "source_caption"
    },
    {
      "start": 381.866,
      "end": 383.1,
      "text": "刚才那个4B模型",
      "source": "source_caption"
    },
    {
      "start": 383.166,
      "end": 384.866,
      "text": "那其他的模型你要下载也是一样的",
      "source": "source_caption"
    },
    {
      "start": 384.866,
      "end": 386.666,
      "text": "在这个地方点下载按钮就好了",
      "source": "source_caption"
    },
    {
      "start": 386.666,
      "end": 387.866,
      "text": "很简单",
      "source": "source_caption"
    },
    {
      "start": 388.0,
      "end": 389.933,
      "text": "然后最后一个这个模型就是",
      "source": "source_caption"
    },
    {
      "start": 389.933,
      "end": 391.7,
      "text": "那个语音合成的这个模型",
      "source": "source_caption"
    },
    {
      "start": 391.7,
      "end": 393.666,
      "text": "同样还是在这里点浏览模型库",
      "source": "source_caption"
    },
    {
      "start": 394.3,
      "end": 395.933,
      "text": "可以看到这底下有很多待选的",
      "source": "source_caption"
    },
    {
      "start": 395.933,
      "end": 397.333,
      "text": "你根据自己的需求来选",
      "source": "source_caption"
    },
    {
      "start": 397.4,
      "end": 398.8,
      "text": "那它这边会有一些标签",
      "source": "source_caption"
    },
    {
      "start": 398.8,
      "end": 401.1,
      "text": "就是什么en, fa这些",
      "source": "source_caption"
    },
    {
      "start": 401.1,
      "end": 403.566,
      "text": "这些是代表你这个语音的模型",
      "source": "source_caption"
    },
    {
      "start": 403.566,
      "end": 404.933,
      "text": "它支持哪些语种",
      "source": "source_caption"
    },
    {
      "start": 404.933,
      "end": 406.3,
      "text": "我这里选择一个很多的",
      "source": "source_caption"
    },
    {
      "start": 406.3,
      "end": 407.133,
      "text": "大家可以看到",
      "source": "source_caption"
    },
    {
      "start": 407.133,
      "end": 408.466,
      "text": "我把它也下载好了",
      "source": "source_caption"
    },
    {
      "start": 408.566,
      "end": 409.3,
      "text": "下载好之后",
      "source": "source_caption"
    },
    {
      "start": 409.3,
      "end": 411.166,
      "text": "这三个模型都配好了OK",
      "source": "source_caption"
    },
    {
      "start": 411.166,
      "end": 413.133,
      "text": "那我们现在要把它选中了",
      "source": "source_caption"
    },
    {
      "start": 413.133,
      "end": 414.766,
      "text": "就是你已经把它都选好了",
      "source": "source_caption"
    },
    {
      "start": 414.766,
      "end": 417.0,
      "text": "然后这里有语音合成的一个速度",
      "source": "source_caption"
    },
    {
      "start": 417.2,
      "end": 417.766,
      "text": "这个东西",
      "source": "source_caption"
    },
    {
      "start": 417.766,
      "end": 418.733,
      "text": "它默认是1.0",
      "source": "source_caption"
    },
    {
      "start": 418.733,
      "end": 420.133,
      "text": "我这里给它改成了1.2",
      "source": "source_caption"
    },
    {
      "start": 420.133,
      "end": 422.266,
      "text": "可能我觉得稍微速度会快一点",
      "source": "source_caption"
    },
    {
      "start": 422.266,
      "end": 424.3,
      "text": "这个根据个人的一个需求吧",
      "source": "source_caption"
    },
    {
      "start": 424.3,
      "end": 425.966,
      "text": "都没有一个绝对的标准",
      "source": "source_caption"
    },
    {
      "start": 426.1,
      "end": 427.5,
      "text": "然后底下这几个参数",
      "source": "source_caption"
    },
    {
      "start": 427.5,
      "end": 428.9,
      "text": "我要给大家稍微提一下",
      "source": "source_caption"
    },
    {
      "start": 428.933,
      "end": 430.8,
      "text": "这些都是默认已经配好的",
      "source": "source_caption"
    },
    {
      "start": 430.8,
      "end": 432.3,
      "text": "它都是一个比较成熟的参数",
      "source": "source_caption"
    },
    {
      "start": 432.3,
      "end": 433.966,
      "text": "你不用一般不用去改的",
      "source": "source_caption"
    },
    {
      "start": 433.966,
      "end": 434.966,
      "text": "像这个什么",
      "source": "source_caption"
    },
    {
      "start": 434.966,
      "end": 437.166,
      "text": "比如说这个最小的静音时长",
      "source": "source_caption"
    },
    {
      "start": 437.6,
      "end": 439.5,
      "text": "这一项它一般是设置为1.4秒",
      "source": "source_caption"
    },
    {
      "start": 439.5,
      "end": 440.266,
      "text": "是什么意思",
      "source": "source_caption"
    },
    {
      "start": 440.266,
      "end": 441.3,
      "text": "就是我们说话的时候",
      "source": "source_caption"
    },
    {
      "start": 441.3,
      "end": 443.866,
      "text": "如果两句话间隔超过了1.4秒的话",
      "source": "source_caption"
    },
    {
      "start": 443.866,
      "end": 446.7,
      "text": "那么它就会把这里切割开",
      "source": "source_caption"
    },
    {
      "start": 446.7,
      "end": 447.566,
      "text": "分成两句话",
      "source": "source_caption"
    },
    {
      "start": 447.566,
      "end": 448.5,
      "text": "就这意思",
      "source": "source_caption"
    },
    {
      "start": 448.533,
      "end": 449.866,
      "text": "然后最小的语音时长",
      "source": "source_caption"
    },
    {
      "start": 449.866,
      "end": 452.0,
      "text": "它这里默认写的是0.4秒",
      "source": "source_caption"
    },
    {
      "start": 452.3,
      "end": 453.0,
      "text": "也就是说",
      "source": "source_caption"
    },
    {
      "start": 453.0,
      "end": 454.133,
      "text": "如果你说话的有一些",
      "source": "source_caption"
    },
    {
      "start": 454.133,
      "end": 455.933,
      "text": "那种声音比较零碎",
      "source": "source_caption"
    },
    {
      "start": 455.933,
      "end": 457.733,
      "text": "那个声音的片段小于0.4秒",
      "source": "source_caption"
    },
    {
      "start": 457.733,
      "end": 458.966,
      "text": "他就不会给你去识别",
      "source": "source_caption"
    },
    {
      "start": 458.966,
      "end": 460.4,
      "text": "就相当于是杂音对吧",
      "source": "source_caption"
    },
    {
      "start": 460.4,
      "end": 462.333,
      "text": "这也是比较节省算力的一种方法",
      "source": "source_caption"
    },
    {
      "start": 462.533,
      "end": 463.866,
      "text": "OK那这个都弄好之后",
      "source": "source_caption"
    },
    {
      "start": 463.866,
      "end": 465.266,
      "text": "我们这里还有一个音频",
      "source": "source_caption"
    },
    {
      "start": 465.466,
      "end": 467.066,
      "text": "这一项也是你的",
      "source": "source_caption"
    },
    {
      "start": 467.2,
      "end": 468.466,
      "text": "就是你的物理的外设",
      "source": "source_caption"
    },
    {
      "start": 468.466,
      "end": 469.333,
      "text": "我们要选什么",
      "source": "source_caption"
    },
    {
      "start": 469.333,
      "end": 470.533,
      "text": "首先你的麦克风",
      "source": "source_caption"
    },
    {
      "start": 470.533,
      "end": 472.666,
      "text": "你要选你的实际的物理的外设",
      "source": "source_caption"
    },
    {
      "start": 472.666,
      "end": 473.6,
      "text": "你不要选其他的",
      "source": "source_caption"
    },
    {
      "start": 473.6,
      "end": 475.2,
      "text": "不然他到时候连不通的",
      "source": "source_caption"
    },
    {
      "start": 475.3,
      "end": 476.5,
      "text": "然后",
      "source": "source_caption"
    },
    {
      "start": 476.5,
      "end": 478.4,
      "text": "这个东西其他一般就不用管了",
      "source": "source_caption"
    },
    {
      "start": 478.4,
      "end": 480.166,
      "text": "然后还有这个供提供商",
      "source": "source_caption"
    },
    {
      "start": 480.166,
      "end": 481.7,
      "text": "这个刚刚我们已经说过了",
      "source": "source_caption"
    },
    {
      "start": 481.733,
      "end": 483.7,
      "text": "对然后就是常规这里",
      "source": "source_caption"
    },
    {
      "start": 483.7,
      "end": 484.9,
      "text": "然后你要说什么",
      "source": "source_caption"
    },
    {
      "start": 484.9,
      "end": 485.866,
      "text": "你要说中文",
      "source": "source_caption"
    },
    {
      "start": 485.966,
      "end": 487.933,
      "text": "然后你希望对方听到什么",
      "source": "source_caption"
    },
    {
      "start": 487.933,
      "end": 488.866,
      "text": "我选英文",
      "source": "source_caption"
    },
    {
      "start": 488.9,
      "end": 490.933,
      "text": "那你这里的语言你要注意",
      "source": "source_caption"
    },
    {
      "start": 491.066,
      "end": 492.566,
      "text": "因为我刚才选的那些模型",
      "source": "source_caption"
    },
    {
      "start": 492.566,
      "end": 493.933,
      "text": "它支持了很多种语种",
      "source": "source_caption"
    },
    {
      "start": 493.933,
      "end": 495.733,
      "text": "所以我这里才有的这么多选",
      "source": "source_caption"
    },
    {
      "start": 496.3,
      "end": 497.666,
      "text": "如果你选的模型它就",
      "source": "source_caption"
    },
    {
      "start": 497.666,
      "end": 498.733,
      "text": "支持一两种语种的话",
      "source": "source_caption"
    },
    {
      "start": 498.733,
      "end": 500.4,
      "text": "你这个下拉框是没这么多选的",
      "source": "source_caption"
    },
    {
      "start": 500.566,
      "end": 501.866,
      "text": "那我们这边选好之后",
      "source": "source_caption"
    },
    {
      "start": 501.866,
      "end": 504.533,
      "text": "我们就点击这里开始会话",
      "source": "source_caption"
    },
    {
      "start": 504.533,
      "end": 505.533,
      "text": "然后左边",
      "source": "source_caption"
    },
    {
      "start": 505.533,
      "end": 506.4,
      "text": "它是有3个",
      "source": "source_caption"
    },
    {
      "start": 506.4,
      "end": 507.466,
      "text": "这里是我",
      "source": "source_caption"
    },
    {
      "start": 507.466,
      "end": 508.533,
      "text": "还有一个对方",
      "source": "source_caption"
    },
    {
      "start": 508.533,
      "end": 509.666,
      "text": "还有一个两者",
      "source": "source_caption"
    },
    {
      "start": 509.766,
      "end": 511.466,
      "text": "那我们目前主要是用",
      "source": "source_caption"
    },
    {
      "start": 511.466,
      "end": 512.466,
      "text": "我这种方式就好了",
      "source": "source_caption"
    },
    {
      "start": 512.666,
      "end": 513.6,
      "text": "如果是对方的话",
      "source": "source_caption"
    },
    {
      "start": 513.6,
      "end": 514.366,
      "text": "那就是另外一种",
      "source": "source_caption"
    },
    {
      "start": 514.366,
      "end": 516.6,
      "text": "这后面这两个我们今天不演示了OK",
      "source": "source_caption"
    },
    {
      "start": 516.666,
      "end": 518.066,
      "text": "然后我们点击开始会话",
      "source": "source_caption"
    },
    {
      "start": 518.566,
      "end": 519.6,
      "text": "这里加载好了之后",
      "source": "source_caption"
    },
    {
      "start": 519.6,
      "end": 520.9,
      "text": "我们现在就可以说话了",
      "source": "source_caption"
    },
    {
      "start": 520.9,
      "end": 522.266,
      "text": "然后你可以看到左上角",
      "source": "source_caption"
    },
    {
      "start": 522.266,
      "end": 525.0,
      "text": "它正在识别我们讲的话",
      "source": "source_caption"
    },
    {
      "start": 525.0,
      "end": 526.6,
      "text": "然后它这个是实时的",
      "source": "source_caption"
    },
    {
      "start": 526.6,
      "end": 527.466,
      "text": "是流式的",
      "source": "source_caption"
    },
    {
      "start": 527.466,
      "end": 528.666,
      "text": "那这个字有点小",
      "source": "source_caption"
    },
    {
      "start": 528.666,
      "end": 529.933,
      "text": "我们可以点一下右上角",
      "source": "source_caption"
    },
    {
      "start": 529.933,
      "end": 531.1,
      "text": "这里有一个字幕",
      "source": "source_caption"
    },
    {
      "start": 531.1,
      "end": 532.166,
      "text": "把它放大一些",
      "source": "source_caption"
    },
    {
      "start": 551.0,
      "end": 551.966,
      "text": "对整体看上去",
      "source": "source_caption"
    },
    {
      "start": 551.966,
      "end": 552.966,
      "text": "它这个效果还可以",
      "source": "source_caption"
    },
    {
      "start": 552.966,
      "end": 553.966,
      "text": "虽然说有时候可能",
      "source": "source_caption"
    },
    {
      "start": 553.966,
      "end": 555.533,
      "text": "识别的个别字不太精准",
      "source": "source_caption"
    },
    {
      "start": 555.533,
      "end": 557.066,
      "text": "那当然这可能也是我个人",
      "source": "source_caption"
    },
    {
      "start": 557.066,
      "end": 558.533,
      "text": "有时候普通话不太标准",
      "source": "source_caption"
    },
    {
      "start": 558.533,
      "end": 559.6,
      "text": "所以它识别起来",
      "source": "source_caption"
    },
    {
      "start": 559.6,
      "end": 561.5,
      "text": "可能会有一些点误差",
      "source": "source_caption"
    },
    {
      "start": 561.5,
      "end": 563.0,
      "text": "但是也不影响它的一个翻译",
      "source": "source_caption"
    },
    {
      "start": 563.0,
      "end": 564.933,
      "text": "我们可以看到它整体",
      "source": "source_caption"
    },
    {
      "start": 564.933,
      "end": 566.5,
      "text": "我觉得效果还是不错的",
      "source": "source_caption"
    },
    {
      "start": 591.933,
      "end": 592.8,
      "text": "OK我们退出的话",
      "source": "source_caption"
    },
    {
      "start": 592.8,
      "end": 593.966,
      "text": "我们就点一下ESC",
      "source": "source_caption"
    },
    {
      "start": 594.266,
      "end": 595.433,
      "text": "然后这些字幕的话",
      "source": "source_caption"
    },
    {
      "start": 595.5,
      "end": 598.366,
      "text": "它有一个删除的小按钮点一下",
      "source": "source_caption"
    },
    {
      "start": 598.466,
      "end": 600.166,
      "text": "它就给它清空了",
      "source": "source_caption"
    },
    {
      "start": 600.166,
      "end": 601.066,
      "text": "然后我们要结束的话",
      "source": "source_caption"
    },
    {
      "start": 601.066,
      "end": 603.2,
      "text": "我们点击下方的这个结束会话",
      "source": "source_caption"
    },
    {
      "start": 603.333,
      "end": 604.6,
      "text": "那这个就结束了",
      "source": "source_caption"
    },
    {
      "start": 604.9,
      "end": 607.4,
      "text": "OK那如果你希望对方听到的",
      "source": "source_caption"
    },
    {
      "start": 607.4,
      "end": 609.533,
      "text": "就我刚刚给大家演示了不同的语种",
      "source": "source_caption"
    },
    {
      "start": 609.533,
      "end": 610.6,
      "text": "这个地方比如说你",
      "source": "source_caption"
    },
    {
      "start": 610.6,
      "end": 612.3,
      "text": "想要他听到西班牙语",
      "source": "source_caption"
    },
    {
      "start": 612.3,
      "end": 614.166,
      "text": "或者是你想要他听到日本语",
      "source": "source_caption"
    },
    {
      "start": 614.166,
      "end": 615.866,
      "text": "那你这个地方切换一下就好了",
      "source": "source_caption"
    },
    {
      "start": 615.866,
      "end": 617.666,
      "text": "然后你这里点击开始会话",
      "source": "source_caption"
    },
    {
      "start": 618.666,
      "end": 620.533,
      "text": "你好可以听到我声音吗",
      "source": "source_caption"
    },
    {
      "start": 625.266,
      "end": 625.733,
      "text": "你好",
      "source": "source_caption"
    },
    {
      "start": 625.733,
      "end": 628.533,
      "text": "我想现在跟你分享一些事情",
      "source": "source_caption"
    },
    {
      "start": 633.9,
      "end": 634.733,
      "text": "你可以看到",
      "source": "source_caption"
    },
    {
      "start": 634.733,
      "end": 635.766,
      "text": "他这个字幕还是",
      "source": "source_caption"
    },
    {
      "start": 635.766,
      "end": 637.733,
      "text": "就是逐步的播过去的",
      "source": "source_caption"
    },
    {
      "start": 637.733,
      "end": 640.1,
      "text": "这个我觉得效果是非常好的",
      "source": "source_caption"
    },
    {
      "start": 640.1,
      "end": 642.333,
      "text": "就是有你讲话",
      "source": "source_caption"
    },
    {
      "start": 642.333,
      "end": 645.5,
      "text": "然后你可以看到这个字幕是在流动的",
      "source": "source_caption"
    },
    {
      "start": 663.766,
      "end": 665.066,
      "text": "OK然后我们把它结束",
      "source": "source_caption"
    },
    {
      "start": 665.066,
      "end": 667.766,
      "text": "然后我们可以再切换其他的这种语言",
      "source": "source_caption"
    },
    {
      "start": 667.766,
      "end": 668.8,
      "text": "那这些都是类似的",
      "source": "source_caption"
    },
    {
      "start": 668.8,
      "end": 670.8,
      "text": "我就不给大家一一去演示了",
      "source": "source_caption"
    },
    {
      "start": 670.966,
      "end": 673.133,
      "text": "然后这上面有一些东西",
      "source": "source_caption"
    },
    {
      "start": 673.133,
      "end": 674.1,
      "text": "我再给大家讲一下",
      "source": "source_caption"
    },
    {
      "start": 674.1,
      "end": 675.5,
      "text": "就是提供商",
      "source": "source_caption"
    },
    {
      "start": 675.733,
      "end": 678.633,
      "text": "那因为本地的其实我已经试过了",
      "source": "source_caption"
    },
    {
      "start": 678.733,
      "end": 680.466,
      "text": "这几个模型它都是在本地",
      "source": "source_caption"
    },
    {
      "start": 680.466,
      "end": 682.766,
      "text": "就没有去走任何在线的API",
      "source": "source_caption"
    },
    {
      "start": 682.766,
      "end": 684.066,
      "text": "那如果有些小伙伴他说",
      "source": "source_caption"
    },
    {
      "start": 684.066,
      "end": 685.566,
      "text": "我有在线的这种API的话",
      "source": "source_caption"
    },
    {
      "start": 685.566,
      "end": 686.4,
      "text": "你也可以去用",
      "source": "source_caption"
    },
    {
      "start": 686.4,
      "end": 687.066,
      "text": "这个都行",
      "source": "source_caption"
    },
    {
      "start": 687.066,
      "end": 688.4,
      "text": "因为我觉得其实本地",
      "source": "source_caption"
    },
    {
      "start": 688.4,
      "end": 689.5,
      "text": "的效果已经很棒了",
      "source": "source_caption"
    },
    {
      "start": 689.5,
      "end": 690.366,
      "text": "大家可以看到",
      "source": "source_caption"
    },
    {
      "start": 690.466,
      "end": 691.566,
      "text": "那这边怎么选",
      "source": "source_caption"
    },
    {
      "start": 691.566,
      "end": 692.3,
      "text": "你可以看到",
      "source": "source_caption"
    },
    {
      "start": 692.3,
      "end": 693.966,
      "text": "首先在这个地方点常规",
      "source": "source_caption"
    },
    {
      "start": 694.266,
      "end": 696.1,
      "text": "然后这里有个AI服务提供商",
      "source": "source_caption"
    },
    {
      "start": 696.1,
      "end": 697.1,
      "text": "我们点一下",
      "source": "source_caption"
    },
    {
      "start": 697.1,
      "end": 698.4,
      "text": "这里有Openai",
      "source": "source_caption"
    },
    {
      "start": 698.466,
      "end": 699.766,
      "text": "有一些其他的",
      "source": "source_caption"
    },
    {
      "start": 699.766,
      "end": 701.1,
      "text": "比如说什么豆包",
      "source": "source_caption"
    },
    {
      "start": 701.1,
      "end": 702.8,
      "text": "还有那个谷歌Gemini呀",
      "source": "source_caption"
    },
    {
      "start": 702.8,
      "end": 703.766,
      "text": "等等的这些",
      "source": "source_caption"
    },
    {
      "start": 703.866,
      "end": 705.166,
      "text": "你有买哪个API",
      "source": "source_caption"
    },
    {
      "start": 705.166,
      "end": 706.8,
      "text": "你就给它对应的输入",
      "source": "source_caption"
    },
    {
      "start": 706.8,
      "end": 708.0,
      "text": "比如我点Openai的",
      "source": "source_caption"
    },
    {
      "start": 708.0,
      "end": 709.333,
      "text": "你看你点进Openai的",
      "source": "source_caption"
    },
    {
      "start": 709.333,
      "end": 710.766,
      "text": "这个地方填入你的密钥",
      "source": "source_caption"
    },
    {
      "start": 710.9,
      "end": 712.4,
      "text": "然后这里音频的话",
      "source": "source_caption"
    },
    {
      "start": 712.4,
      "end": 713.466,
      "text": "这个保持不动",
      "source": "source_caption"
    },
    {
      "start": 713.6,
      "end": 714.733,
      "text": "提供商这一块的话",
      "source": "source_caption"
    },
    {
      "start": 714.733,
      "end": 715.5,
      "text": "你可以选",
      "source": "source_caption"
    },
    {
      "start": 715.5,
      "end": 716.5,
      "text": "在这里选择",
      "source": "source_caption"
    },
    {
      "start": 716.5,
      "end": 718.733,
      "text": "就是Openai那个语音的模型对应的",
      "source": "source_caption"
    },
    {
      "start": 718.733,
      "end": 721.0,
      "text": "发音的那个角色对吧",
      "source": "source_caption"
    },
    {
      "start": 721.0,
      "end": 721.866,
      "text": "你自己去选",
      "source": "source_caption"
    },
    {
      "start": 721.866,
      "end": 723.3,
      "text": "然后这些东西你可以把",
      "source": "source_caption"
    },
    {
      "start": 723.3,
      "end": 724.8,
      "text": "它参数对应的填过去",
      "source": "source_caption"
    },
    {
      "start": 724.933,
      "end": 727.066,
      "text": "它就可以调在线的API了，对",
      "source": "source_caption"
    },
    {
      "start": 727.133,
      "end": 728.333,
      "text": "那原理是一样的",
      "source": "source_caption"
    },
    {
      "start": 728.333,
      "end": 729.4,
      "text": "但是我自己是觉得",
      "source": "source_caption"
    },
    {
      "start": 729.4,
      "end": 731.3,
      "text": "就是其实我们免费的",
      "source": "source_caption"
    },
    {
      "start": 731.3,
      "end": 732.166,
      "text": "就是我今天给大家",
      "source": "source_caption"
    },
    {
      "start": 732.166,
      "end": 733.5,
      "text": "展示的主要是免费的",
      "source": "source_caption"
    },
    {
      "start": 733.766,
      "end": 735.3,
      "text": "已经够用了对",
      "source": "source_caption"
    },
    {
      "start": 735.333,
      "end": 737.7,
      "text": "所以整体实现下来这个效果是",
      "source": "source_caption"
    },
    {
      "start": 737.7,
      "end": 738.333,
      "text": "我觉得是在我目前",
      "source": "source_caption"
    },
    {
      "start": 739.4,
      "end": 740.866,
      "text": "就是看过的所有的开源的",
      "source": "source_caption"
    },
    {
      "start": 740.866,
      "end": 741.366,
      "text": "这种同声传译的项目里面",
      "source": "source_caption"
    },
    {
      "start": 742.1,
      "end": 743.033,
      "text": "应该是最好的了",
      "source": "source_caption"
    },
    {
      "start": 743.033,
      "end": 744.4,
      "text": "所以我强烈建议大家",
      "source": "source_caption"
    },
    {
      "start": 744.4,
      "end": 747.133,
      "text": "尤其是很多小伙伴经常需要",
      "source": "source_caption"
    },
    {
      "start": 747.266,
      "end": 748.8,
      "text": "就是用到外语去交流",
      "source": "source_caption"
    },
    {
      "start": 748.8,
      "end": 749.8,
      "text": "但是自己可能外语",
      "source": "source_caption"
    },
    {
      "start": 749.8,
      "end": 751.166,
      "text": "水平可能还没那么好",
      "source": "source_caption"
    },
    {
      "start": 751.166,
      "end": 752.366,
      "text": "用这些工具我觉得是",
      "source": "source_caption"
    },
    {
      "start": 752.366,
      "end": 754.066,
      "text": "对你有很大的帮助的OK",
      "source": "source_caption"
    },
    {
      "start": 754.266,
      "end": 756.6,
      "text": "那本期的这些课程资源其实比较少",
      "source": "source_caption"
    },
    {
      "start": 756.6,
      "end": 757.733,
      "text": "但是我会放到评论区",
      "source": "source_caption"
    },
    {
      "start": 757.733,
      "end": 758.933,
      "text": "大家直接可以去用",
      "source": "source_caption"
    },
    {
      "start": 758.933,
      "end": 761.1,
      "text": "那我最后还要提到一个注意点",
      "source": "source_caption"
    },
    {
      "start": 761.1,
      "end": 762.766,
      "text": "就是我在实际使用过程中",
      "source": "source_caption"
    },
    {
      "start": 763.366,
      "end": 765.2,
      "text": "很多小伙伴使用的时候会报错",
      "source": "source_caption"
    },
    {
      "start": 765.333,
      "end": 767.133,
      "text": "那这个报错是什么原因造成的",
      "source": "source_caption"
    },
    {
      "start": 767.133,
      "end": 769.166,
      "text": "就是我们一定要注意",
      "source": "source_caption"
    },
    {
      "start": 769.266,
      "end": 771.566,
      "text": "我们用到的这三个模型",
      "source": "source_caption"
    },
    {
      "start": 771.566,
      "end": 773.333,
      "text": "它支持的语种范围是什么",
      "source": "source_caption"
    },
    {
      "start": 773.966,
      "end": 774.666,
      "text": "举个例子",
      "source": "source_caption"
    },
    {
      "start": 774.666,
      "end": 776.366,
      "text": "比如说我这里的语种",
      "source": "source_caption"
    },
    {
      "start": 776.366,
      "end": 776.866,
      "text": "它有一个粤语",
      "source": "source_caption"
    },
    {
      "start": 777.5,
      "end": 778.333,
      "text": "这也是支持的",
      "source": "source_caption"
    },
    {
      "start": 778.4,
      "end": 781.166,
      "text": "我们可以看到这有粤语对吧但是",
      "source": "source_caption"
    },
    {
      "start": 781.333,
      "end": 782.366,
      "text": "我这里说话的话",
      "source": "source_caption"
    },
    {
      "start": 782.366,
      "end": 784.0,
      "text": "我说的是普通话然后",
      "source": "source_caption"
    },
    {
      "start": 784.0,
      "end": 786.0,
      "text": "我这个语音识别可能它这个模型",
      "source": "source_caption"
    },
    {
      "start": 786.333,
      "end": 787.733,
      "text": "它能够识别普通话",
      "source": "source_caption"
    },
    {
      "start": 787.733,
      "end": 790.266,
      "text": "但是它语音合成的那个模型",
      "source": "source_caption"
    },
    {
      "start": 790.266,
      "end": 792.0,
      "text": "它不支持播放粤语",
      "source": "source_caption"
    },
    {
      "start": 792.0,
      "end": 793.066,
      "text": "那么它就会报错",
      "source": "source_caption"
    },
    {
      "start": 793.066,
      "end": 794.166,
      "text": "你这三个模型",
      "source": "source_caption"
    },
    {
      "start": 795.1,
      "end": 796.8,
      "text": "大家可以看到",
      "source": "source_caption"
    },
    {
      "start": 796.8,
      "end": 798.133,
      "text": "就是你这三个模型",
      "source": "source_caption"
    },
    {
      "start": 798.2,
      "end": 800.9,
      "text": "它们支持的语种一定要有共同的",
      "source": "source_caption"
    },
    {
      "start": 801.4,
      "end": 803.0,
      "text": "那么这个链路才是通的",
      "source": "source_caption"
    },
    {
      "start": 803.4,
      "end": 804.733,
      "text": "有一个不支持的",
      "source": "source_caption"
    },
    {
      "start": 804.733,
      "end": 806.233,
      "text": "那么你这个链路就通不了",
      "source": "source_caption"
    },
    {
      "start": 806.466,
      "end": 807.166,
      "text": "小伙伴可能遇到的这些",
      "source": "source_caption"
    },
    {
      "start": 807.733,
      "end": 809.9,
      "text": "大部分可能是这一方面的问题",
      "source": "source_caption"
    },
    {
      "start": 809.9,
      "end": 811.6,
      "text": "还有就是你网络不好",
      "source": "source_caption"
    },
    {
      "start": 811.6,
      "end": 813.1,
      "text": "你网络如果有问题的话",
      "source": "source_caption"
    },
    {
      "start": 813.1,
      "end": 814.2,
      "text": "你没开全局模式的话",
      "source": "source_caption"
    },
    {
      "start": 814.2,
      "end": 816.1,
      "text": "你这些模型是下载不到你本地的",
      "source": "source_caption"
    },
    {
      "start": 816.1,
      "end": 816.966,
      "text": "还有一个细节就是",
      "source": "source_caption"
    },
    {
      "start": 816.966,
      "end": 818.166,
      "text": "关于这里有个登录",
      "source": "source_caption"
    },
    {
      "start": 818.166,
      "end": 819.9,
      "text": "这里刚才跟大家有讲过",
      "source": "source_caption"
    },
    {
      "start": 819.9,
      "end": 821.066,
      "text": "你看这个登录这里",
      "source": "source_caption"
    },
    {
      "start": 821.066,
      "end": 822.133,
      "text": "你可以点击去注册",
      "source": "source_caption"
    },
    {
      "start": 822.133,
      "end": 824.7,
      "text": "或者是就已经有账号就登录进去",
      "source": "source_caption"
    },
    {
      "start": 824.9,
      "end": 825.766,
      "text": "那它本质上",
      "source": "source_caption"
    },
    {
      "start": 825.766,
      "end": 828.066,
      "text": "它会用到它云端的这个sokuji",
      "source": "source_caption"
    },
    {
      "start": 828.066,
      "end": 829.3,
      "text": "它自己的翻译服务 api",
      "source": "source_caption"
    },
    {
      "start": 829.6,
      "end": 832.0,
      "text": "那对于有些小伙伴不想去折腾",
      "source": "source_caption"
    },
    {
      "start": 832.0,
      "end": 832.9,
      "text": "你可以去用它",
      "source": "source_caption"
    },
    {
      "start": 832.9,
      "end": 833.933,
      "text": "那这样的话你就不用",
      "source": "source_caption"
    },
    {
      "start": 833.933,
      "end": 835.266,
      "text": "去本地下载模型了",
      "source": "source_caption"
    },
    {
      "start": 835.266,
      "end": 836.966,
      "text": "但是要提醒一下大家就是",
      "source": "source_caption"
    },
    {
      "start": 836.966,
      "end": 837.666,
      "text": "你这样的话",
      "source": "source_caption"
    },
    {
      "start": 837.666,
      "end": 839.533,
      "text": "你的数据是会经过它的服务器的",
      "source": "source_caption"
    },
    {
      "start": 839.533,
      "end": 840.966,
      "text": "就是说数据的隐私性",
      "source": "source_caption"
    },
    {
      "start": 840.966,
      "end": 841.466,
      "text": "就保障没那么好",
      "source": "source_caption"
    },
    {
      "start": 841.866,
      "end": 842.866,
      "text": "根据自己的需求来",
      "source": "source_caption"
    },
    {
      "start": 842.866,
      "end": 844.333,
      "text": "如果你全都想让数据不",
      "source": "source_caption"
    },
    {
      "start": 844.333,
      "end": 845.7,
      "text": "出自己的本地的电脑",
      "source": "source_caption"
    },
    {
      "start": 845.7,
      "end": 847.566,
      "text": "那么你就用我刚刚给大家讲的方法",
      "source": "source_caption"
    },
    {
      "start": 847.7,
      "end": 848.333,
      "text": "还有一点",
      "source": "source_caption"
    },
    {
      "start": 848.333,
      "end": 849.866,
      "text": "也是我个人使用的一个经验",
      "source": "source_caption"
    },
    {
      "start": 849.866,
      "end": 852.133,
      "text": "这个东西对大家应该很有帮助",
      "source": "source_caption"
    },
    {
      "start": 852.533,
      "end": 853.7,
      "text": "同声传译这个东西",
      "source": "source_caption"
    },
    {
      "start": 853.7,
      "end": 855.8,
      "text": "大家要知道它最看重的是什么",
      "source": "source_caption"
    },
    {
      "start": 855.8,
      "end": 858.133,
      "text": "就是一个是翻译的准确性",
      "source": "source_caption"
    },
    {
      "start": 858.166,
      "end": 859.6,
      "text": "一个是它的一个时延",
      "source": "source_caption"
    },
    {
      "start": 859.733,
      "end": 861.266,
      "text": "就是我们那个翻译的",
      "source": "source_caption"
    },
    {
      "start": 861.266,
      "end": 862.566,
      "text": "那个时间不能太久了",
      "source": "source_caption"
    },
    {
      "start": 862.566,
      "end": 864.2,
      "text": "这样的话就做不到同声了",
      "source": "source_caption"
    },
    {
      "start": 864.2,
      "end": 866.933,
      "text": "那我们要想在这个项目中",
      "source": "source_caption"
    },
    {
      "start": 866.933,
      "end": 869.666,
      "text": "要尽量的去把它时延降低",
      "source": "source_caption"
    },
    {
      "start": 869.666,
      "end": 871.266,
      "text": "让它的用户体验更好",
      "source": "source_caption"
    },
    {
      "start": 871.266,
      "end": 872.466,
      "text": "应该怎么去做",
      "source": "source_caption"
    },
    {
      "start": 872.5,
      "end": 874.1,
      "text": "首先刚才我跟大家是讲了",
      "source": "source_caption"
    },
    {
      "start": 874.1,
      "end": 875.6,
      "text": "它里面有很多模型对吧",
      "source": "source_caption"
    },
    {
      "start": 875.6,
      "end": 878.2,
      "text": "大家可以看到这模型是有一大堆的",
      "source": "source_caption"
    },
    {
      "start": 878.366,
      "end": 880.2,
      "text": "那么我们应该怎么去选模型",
      "source": "source_caption"
    },
    {
      "start": 880.2,
      "end": 881.1,
      "text": "大家要记住",
      "source": "source_caption"
    },
    {
      "start": 881.3,
      "end": 883.366,
      "text": "你要选肯定要去选流式的",
      "source": "source_caption"
    },
    {
      "start": 883.6,
      "end": 885.6,
      "text": "也就是说不要选非流式的模型",
      "source": "source_caption"
    },
    {
      "start": 885.6,
      "end": 887.066,
      "text": "那流式的模型一般会带上",
      "source": "source_caption"
    },
    {
      "start": 887.066,
      "end": 889.0,
      "text": "一个叫realtime的这个字段",
      "source": "source_caption"
    },
    {
      "start": 889.0,
      "end": 891.7,
      "text": "然后有这个Web GPU的",
      "source": "source_caption"
    },
    {
      "start": 891.7,
      "end": 893.5,
      "text": "你就尽量去选它对",
      "source": "source_caption"
    },
    {
      "start": 893.566,
      "end": 894.966,
      "text": "这个是一个细节",
      "source": "source_caption"
    },
    {
      "start": 894.966,
      "end": 895.566,
      "text": "很重要的",
      "source": "source_caption"
    },
    {
      "start": 895.566,
      "end": 897.266,
      "text": "就是你选那种非流式的",
      "source": "source_caption"
    },
    {
      "start": 897.266,
      "end": 898.1,
      "text": "它推你的话",
      "source": "source_caption"
    },
    {
      "start": 898.1,
      "end": 899.5,
      "text": "那整个链路就会很慢了",
      "source": "source_caption"
    },
    {
      "start": 899.5,
      "end": 902.266,
      "text": "等于说你说完半句话以后",
      "source": "source_caption"
    },
    {
      "start": 902.333,
      "end": 904.133,
      "text": "别人都已经说完下一句了",
      "source": "source_caption"
    },
    {
      "start": 904.133,
      "end": 905.266,
      "text": "它还没翻译完",
      "source": "source_caption"
    },
    {
      "start": 905.266,
      "end": 907.466,
      "text": "那当然那那体验就很不好了对吧",
      "source": "source_caption"
    },
    {
      "start": 907.466,
      "end": 908.766,
      "text": "所以大家一定要记住",
      "source": "source_caption"
    },
    {
      "start": 908.766,
      "end": 910.533,
      "text": "就是要尽量选流式的模型",
      "source": "source_caption"
    },
    {
      "start": 910.7,
      "end": 911.966,
      "text": "不管是语音识别",
      "source": "source_caption"
    },
    {
      "start": 911.966,
      "end": 913.1,
      "text": "还是那个翻译",
      "source": "source_caption"
    },
    {
      "start": 913.1,
      "end": 914.466,
      "text": "或者是那个语音合成",
      "source": "source_caption"
    },
    {
      "start": 914.466,
      "end": 915.733,
      "text": "都要去选流式的",
      "source": "source_caption"
    },
    {
      "start": 915.733,
      "end": 917.1,
      "text": "然后有GPU的话",
      "source": "source_caption"
    },
    {
      "start": 917.1,
      "end": 919.8,
      "text": "就要去把那个webgpu那个也选好",
      "source": "source_caption"
    },
    {
      "start": 919.966,
      "end": 920.8,
      "text": "那本期视频",
      "source": "source_caption"
    },
    {
      "start": 920.8,
      "end": 922.2,
      "text": "我就跟大家分享到这了",
      "source": "source_caption"
    },
    {
      "start": 922.2,
      "end": 923.066,
      "text": "大家如果觉得我的",
      "source": "source_caption"
    },
    {
      "start": 923.066,
      "end": 924.533,
      "text": "视频对你有帮助的话",
      "source": "source_caption"
    },
    {
      "start": 924.566,
      "end": 926.0,
      "text": "也请支持一下我的创作",
      "source": "source_caption"
    },
    {
      "start": 926.0,
      "end": 927.5,
      "text": "帮我点赞关注转发",
      "source": "source_caption"
    },
    {
      "start": 927.5,
      "end": 929.066,
      "text": "那本期内容到此结束",
      "source": "source_caption"
    },
    {
      "start": 929.066,
      "end": 930.2,
      "text": "我们下期再会",
      "source": "source_caption"
    }
  ],
  "transcript_selection": {
    "schema": "videonote.transcript_selection.v1",
    "status": "SELECTED",
    "requested_language": "zh",
    "duration": 930.795125,
    "reference_path": "/Users/sagawa/AI_WORK/video_notes/url/20260902_080245__AI同声传译Sokuji_开源免费_多语种_跨平台_Win_Mac_Linux_甚至免GPU_不会外语也能实时交流_Hex-电脑课堂/reference_subtitle.vtt",
    "reference_kind": "manual",
    "reference_language": "zh",
    "selected_source": "youtube_manual",
    "decision_reason": "source_timing_retained",
    "whisper_rejected": false,
    "source_profile": {
      "segment_count": 493,
      "multilingual": false,
      "first_start": 0.533,
      "final_end": 930.2,
      "duration_coverage": 0.999361,
      "maximum_gap_seconds": 25.433,
      "inverted_count": 0,
      "language": {
        "requested": "zh",
        "matched": true,
        "latin": 232,
        "han": 4392,
        "kana": 0,
        "hangul": 0
      },
      "hallucination_region_count": 0,
      "hallucination_regions": [],
      "usable_full": true,
      "usable_partial": true
    },
    "whisper_profile": {
      "segment_count": 523,
      "multilingual": false,
      "first_start": 0.0,
      "final_end": 929.857,
      "duration_coverage": 0.998992,
      "maximum_gap_seconds": 13.917,
      "inverted_count": 0,
      "language": {
        "requested": "zh",
        "matched": true,
        "latin": 540,
        "han": 4302,
        "kana": 17,
        "hangul": 0
      },
      "hallucination_region_count": 0,
      "hallucination_regions": [],
      "usable_full": true,
      "usable_partial": true
    },
    "comparison": {
      "window_seconds": 30.0,
      "compared_window_count": 31,
      "median_score": 0.9276,
      "divergent_window_count": 1,
      "divergent_ratio": 0.0323,
      "divergent_windows": [
        {
          "start": 570.0,
          "end": 600.0,
          "score": 0.3072,
          "sequence_similarity": 0.1985,
          "token_f1": 0.5091,
          "divergent": true
        }
      ]
    },
    "timestamp_alignment": {
      "applied": false,
      "estimated_offset_seconds": 0.0,
      "matched_segment_count": 0,
      "match_scores": []
    },
    "caption_timing_repair": {
      "method": "monotonic_multi_anchor_piecewise_interpolation",
      "applied": false,
      "anchor_count": 0,
      "anchors": [],
      "source_block_count": 63,
      "whisper_block_count": 65,
      "matched_block_count": 27,
      "case_normalization": {
        "applied": false,
        "changed_segment_count": 0
      },
      "source_text_preserved": true,
      "reason": "source_timing_retained"
    },
    "filled_source_profile": {
      "segment_count": 46,
      "multilingual": false,
      "first_start": 0.533,
      "final_end": 930.2,
      "duration_coverage": 0.999361,
      "maximum_gap_seconds": 10.333,
      "inverted_count": 0,
      "language": {
        "requested": "zh",
        "matched": true,
        "latin": 232,
        "han": 4392,
        "kana": 0,
        "hangul": 0
      },
      "hallucination_region_count": 0,
      "hallucination_regions": [],
      "usable_full": true,
      "usable_partial": true
    },
    "whisper_duration_guard": {
      "media_duration": 930.795,
      "rejected_segment_count": 0,
      "rejected_segments": []
    }
  }
}