{
  "text": "M5 Pro去完成了5個維度的本地AI測試。到底它能做些什麼?我們這期把它講清楚。我發現關於MacBook測評卻少之又少。我認為MacBook在本地AI的部署上面還是比較有優勢的。我在我的官網上進行過文章的詳細分析。從硬件能力相關的,比如價格、推理的速度、減存的使用,包括內昏、帶寬、美麗進行了分析。那麼有詳細的數據。感興趣的話可以參考去官網上面相關的文章去看一看。當然我的公眾號上也有同步的文章。\n网上的视频\n然后他们都很少涉及到一个维度\n就是这个噪音\n觉得这个噪音其实挺重要的\n如果你是在家里使用的话\n这个分贝的噪音其实挺繁人的\n在一般人的家里都不会特别大\n可能如果空间小的话\n你总是有这么一个设备\n一只呜呜的声音\n特别吵人的话\n你就算再高的设备\n我觉得有的人觉得可能也没法用\nMacBook Pro呢\n我的一个感觉就是\n它的这个噪音控制是确实惊艳\n第一部分是最常见的\nLLM的文字处理能力\n那么本次使用的模型具体的有四个\n两个GAM4和两个Queen 3.6\n那么它的架构和相关的数据都在这\n发现就是他们的推理速度还是非常可用的\n尤其是GAM4的E4B模型和这个千万三的35B MOE模型\n那么虽然这个E4B是一个Dance\n也就是重蜜结构的模型\n但是呢因为它的每次激活是大概有4B的这个参数\n所以它的推理速度也是相当快\n我之前使用的经验\n我发现在具体使用上\nGAMMA模型它的图形的夜能力\n比千万3.6是高一些\n同时我还测量了所谓的月域能力\n也就是BrokerJag的能力\nBrokerJag应该就是方便生成一些\n容易造成误判的一些内容\n包括像一些踩在边界上的内容\n那么在线模型可能生成起来困难\n这个本地生成起来相对就容易很多\n在具体的测试题目上\n我是用这个Cloud生成了一些测试题目\n然后让他帮着完全全自动测试\n内容特别多\n我就不一个一个展示所有的这个答案了\n然后这是一些具体的数据\n总体给我的感觉是这样的\n它的题目都能比较好的完成预定的效果\n比如说像有分苹果的\n然后分几个的简单的这种盘数\n包括求函数的这种数学题\n或写代码的这种效果\n或修复bug\n包括判断说真假话推理过程\n或解释一些话\n然后用看看具体的执行的能力等等\n就后续的这些答案呢\n我后期将会逐渐整理\n也去把它放到我这个模型测试的\n这个页面的后边供您参考\n第二个我测试的是TTS语音\n语音合成能力\n这里边呢\n我目前搜索的资料是这样的\n它有两种架构\n一个是原声支持MLX\n这个METAL加速功能的这个架构\n那么它的速度稍微快一些\n还有一种是直接用原声这个格式的\n不是支持这个METAL加速的\n那么它俩有一个区别\n原声支持加速的呢\n它在这个进行声音克隆的时候呢\n它是没有那个功能的\n只能支持文字描述声音\n那换句话说就是\n你用一段话去让它生成一个你想要的声音\n它根据了你描述的话去生成\n然后速度还是比较快的\n它因为没有真正的实现具体声音克隆\n所以只能用另一个叫千万三的Base模型\n这个模型是原版的\n它的速度稍微慢一些\n格式呢不是最快的\n有意思发现就是\n直接用Voice Design\n也就是描述生成这个音频功能呢\n它的速度反而还比这个声音可能要慢一些\n不知道这是因为\n为什么\n可能是因为处理架构上的音\n需要注意的是它有一个bug\n就是截取的那个具体参考的参数\n所你截取的音频是需要匹配\n另外一个五秒即可\n如果这个内容特别多\n或者它两者比匹配的话\n会生成一些乱码\n真人的录音\n我们听一听这个录音\n叫做装算赤\n副算黑\n赤就是红色\n我们这次听一下\n它克隆完毕之后的声音的感觉\n大家可以听到\n这个克隆的声音简直是一家论真\n几乎听不出任何的区别\nTTLS语音合成功能上\n我觉得还是相当强大\n第三点\n看图片生成能力\n我这里选用的是质量非常高的\n这个\nFluzy1的DVE的模型\n后侧的是两方面\n一个是直接生成图片\n另一个是使用Laura\n那么具体的生成时间在这里\n风景这个图片生成的时候呢\n它因为需要进行这个冷启动\n然后注入参数\n所以它的生成速度是最慢的\n然后接下来速度就相对来说稳定很多\n大概生成一幅像1024x1024\n也就是它的分辨率相对较高的情况下\n每一幅的图片大概是用到的\n71分到11分\n直接用Laura进行稳定\n这个真人图片克隆的情况下\n也是大概这个相同的速度\n看一下结果\n第一个是风景\n我们看到这些细节给人的感觉还是非常不错的\n效果很好\n赛博朋克的业绩\n专业商务人像\n比较抽象的液态金属\n产品设计度光影\n还是非常不错的\n比较好的执行了我需要的这个要求\n我之前没有任何训练经验\n所以我就自己\n我就自己随意拍了很多照片\n各种角度\n各种的方向\n然后各种场景\n尽量穿不同的衣服\n然后装束尽量去不一样\n然后拍摄了远近各种各样的照片\n数量是60多张\n然后加在一块\n所用的触发词\n这里的触发词就是一个所谓比较生辟的词\n然后帮助触发这个人像\n使用的相关参数是scale等1.0\n是第一张商务头像\n这个可以看起来\n就是样子是不戴眼镜的\n是我本人可能有这个送眉的习惯\n它这个地方呢\n就特别明显的捕捉到\n然后剩下两边\n我这个因为现在正在处于一直控制提供的状态下\n所以胖瘦呢有时候不太一定\n像这张照片就生成的稍微胖一点\n然后这张也是\n这是戴眼镜\n大家看到我俩现在的样子\n就先回来的\n这个是科技演讲\n然后科技演讲的呢\n这个侧面我发现的\n因为可能我给它的样子\n侧面的照片特别多\n所以这个侧面\n我觉得还非常非常像\n然后但是有一个问题\n它出现了比较严重的幻觉\n就是这个手出现了多个\n然后并且胸口\n这有一个我给它的提示词\n就伸透出来了\nHIFEN TAG\n这次生成呢\n我是直接\n不管它生成出什么来\n我全原样展示出来\n所以这是不是经过我选择的\n这就是直接一次性的\n所以等于\n我觉得成功率相对来说\n还是可用的\n非常可用的\n可能得分场景\n然后可能稍微抽抽卡\n一系列的大家看到\n一个人像的样子\n基本上是大概六成吧\n我把四成这个删除了\n六成之后生成出来的样子\n是这样\n视频生成这个部分\n那么大致是两类\n一类是完成视频\n还有一类是图生视频\n然后它的方式呢\n有两类\n一类是蒸馏模式\n速度快\n还有一类是2C\n是两阶段模式\n速度迈起\n它俩之间的速度相差\n比较明显的\n左侧是蒸馏版\n右侧是这个\n所谓更加精细的两阶段版本\n速度相差45倍吧\n两种版本各生成两\n我们先说是第一个\n城市街道这个是蒸油版生成的\n明显描述是让它生成城市街道\n但是它生成了一个室内景\n有车库刚才有白色人物\n跟城市街道的明显偏离了\n所以理解确实有问题\n第二个城市街道呢\n两街道版本\n确实生成的街道有了提升\n然后有人了\n然后有人群了\n有建筑物了\n然后时间也增大了很多\n但是这是夜间警官\n它稍微理解有点问题\n这点是有点缺陷\n第二个是科技实验室\n在科技实验室里边\n这个声唱令利母女人剩下完全的要素\n基本上没有体现出来\n像手啊键盘啊灯光啊实验室这种感觉啊\n後兩階的版本呢,生成了一個這樣的視頻。一個類似於科技人員把它腳化。有點像保安或者快遞啊,外國。\n然後它生成了這些像手臂啊、屏幕這種樣子。接近了這個描述的樣子。\n但是核心动作包括这种感觉是没有的\n我也发现了\n好像fts模型对于生成纯风景\n或者纯那种建筑物之类的那种模式\n他经常愿意加人进去\n就每一个都要有点人\n可能比较富有人性\n这是他的问题\n为了去改变这个问题\n我发现的办法就是用无声视频的方式\n给他一些图\n然后让他去发挥\n那样的话效果会好很多\n无声视频\n然后让图片说话\n找一张随机图库去测试\n可是因为他是自动测试\n我们有干预\n他测试的时候找他\n本来说想说一个人让他说话\n说一个人让他说话\n结果找成了一只狗\n最终生成的样子确实是一只狗\n他没有说最后变成人活还是怎么样的\n但是他这个样子就勉强动动嘴\n如果用真人我之前试着\n本次测试没有试过来\n几秒钟之后\n一两秒立马就变形了\n具体生成的视频的样子\n我们简单看一下\n中流版的\n这是美女\n运动美女\n还得车库的\n是两阶段版本的两个\n很明显动作\n他的样子变得就更加活跃一点\n动作也有了\n是实验室的\n一个阿三在这白话\n就像是一个泰国人他在说话\n你看他脸上这种雀斑这种表情\n我们能看到他确实挺像真人的\n这个效果是不错的\n综合评价我觉得就差很多了\n首先速度相对慢\n然后这个T14理解稍显差\n然后身份保持也不太行\n所以我觉得是这样\n这个视频生成能力\n目前LTX模型\n如果用一些视频里边的匕骤\n或者进行一些原型的\n比如展示一下你想要的东西\n但是直接可能不行\n我试了很多次全完蛋了\n下一部分世界理解部分\n这个世界理解部分呢\n这个模型是这样的\n这个模型带V的\nVision 7B一般的模型\n专门是理解图片\n第一个是代码识别\n第二个是一些图表识别\n第三个是界面\n这是一个模拟的计算机编程界面\n然后还有一个识别图片\n还是依旧是用这个狗为识别的\n本来说是个人\n然后让它识别\n确实最后输出了\n没有人物\n只有一只哈布狗\n这个相当厉害\n不会遵从人的这种强制的复杂图片\n需要识别的时候\n它在统计数字的上面\n确实不太主意\n双图对比的时候\n天鹅宝和在安风景\n他说是跟咖啡馆有关\n有点偏差\n是具体的一些结果\n那么描述比如说场景\n什么样子的\n主体什么样的\n大楼\n包括毛树人物\n这边他说了图片只有一只哈布狗\n有一个分析图表\n然后他给出了具体年月日\n最高值平均值等等等等的\n识别出来了\n然后代码理解\n他理解出来的相关的内容\n贝波那奇数列\n数东西\n然后他数列大堆\n这个数彩的花椒\n分析图片\n然后分析相同点\n不同点\n他说了包括哈维胡\n当然刚才看了\n这是风景\n最终结论吧\n首先实时对话这个是非常可用的\n流畅然后准确\n视觉理解是非常棒的\n大家看到\n它能准确识别数字啊\nC些图柄什么的\n外界的问题不大\n然后声音克隆\n刚才大家听到\n简直是完美\n这个图像生成呢\n等于是速度稍微慢\n克隆震灵来提升\n视觉模型的差距非常大\n只能生成一些基本的内容\n给人用氛围感\n人脸一致性\n基本上不可用\n这个MacBook Pro M564G配置上\n然后小而快的模型\n还是比较有价值\n相信随着苹果架构本身的它的完善\n相关配置环境的完善\n我觉得它的提升应该还是会很迅速的",
  "language": "zh",
  "duration": 614.805313,
  "segments": [
    {
      "start": 0.0,
      "end": 29.9,
      "text": "M5 Pro去完成了5個維度的本地AI測試。到底它能做些什麼?我們這期把它講清楚。我發現關於MacBook測評卻少之又少。我認為MacBook在本地AI的部署上面還是比較有優勢的。我在我的官網上進行過文章的詳細分析。從硬件能力相關的,比如價格、推理的速度、減存的使用,包括內昏、帶寬、美麗進行了分析。那麼有詳細的數據。感興趣的話可以參考去官網上面相關的文章去看一看。當然我的公眾號上也有同步的文章。",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15341398909285262,
      "no_speech_prob": 1.0250230352004674e-11,
      "compression_ratio": 1.355889724310777,
      "source": "whisper"
    },
    {
      "start": 29.9,
      "end": 30.5,
      "text": "网上的视频",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1320807528945635,
      "no_speech_prob": 7.681738578568797e-11,
      "compression_ratio": 1.4581151832460733,
      "source": "whisper"
    },
    {
      "start": 30.5,
      "end": 32.6,
      "text": "然后他们都很少涉及到一个维度",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1320807528945635,
      "no_speech_prob": 7.681738578568797e-11,
      "compression_ratio": 1.4581151832460733,
      "source": "whisper"
    },
    {
      "start": 32.6,
      "end": 33.24,
      "text": "就是这个噪音",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1320807528945635,
      "no_speech_prob": 7.681738578568797e-11,
      "compression_ratio": 1.4581151832460733,
      "source": "whisper"
    },
    {
      "start": 33.24,
      "end": 34.5,
      "text": "觉得这个噪音其实挺重要的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1320807528945635,
      "no_speech_prob": 7.681738578568797e-11,
      "compression_ratio": 1.4581151832460733,
      "source": "whisper"
    },
    {
      "start": 34.5,
      "end": 36.22,
      "text": "如果你是在家里使用的话",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1320807528945635,
      "no_speech_prob": 7.681738578568797e-11,
      "compression_ratio": 1.4581151832460733,
      "source": "whisper"
    },
    {
      "start": 36.22,
      "end": 37.9,
      "text": "这个分贝的噪音其实挺繁人的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1320807528945635,
      "no_speech_prob": 7.681738578568797e-11,
      "compression_ratio": 1.4581151832460733,
      "source": "whisper"
    },
    {
      "start": 37.9,
      "end": 39.28,
      "text": "在一般人的家里都不会特别大",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1320807528945635,
      "no_speech_prob": 7.681738578568797e-11,
      "compression_ratio": 1.4581151832460733,
      "source": "whisper"
    },
    {
      "start": 39.28,
      "end": 40.46,
      "text": "可能如果空间小的话",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1320807528945635,
      "no_speech_prob": 7.681738578568797e-11,
      "compression_ratio": 1.4581151832460733,
      "source": "whisper"
    },
    {
      "start": 40.46,
      "end": 41.76,
      "text": "你总是有这么一个设备",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1320807528945635,
      "no_speech_prob": 7.681738578568797e-11,
      "compression_ratio": 1.4581151832460733,
      "source": "whisper"
    },
    {
      "start": 41.76,
      "end": 42.74,
      "text": "一只呜呜的声音",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1320807528945635,
      "no_speech_prob": 7.681738578568797e-11,
      "compression_ratio": 1.4581151832460733,
      "source": "whisper"
    },
    {
      "start": 42.74,
      "end": 43.9,
      "text": "特别吵人的话",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1320807528945635,
      "no_speech_prob": 7.681738578568797e-11,
      "compression_ratio": 1.4581151832460733,
      "source": "whisper"
    },
    {
      "start": 43.9,
      "end": 45.1,
      "text": "你就算再高的设备",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1320807528945635,
      "no_speech_prob": 7.681738578568797e-11,
      "compression_ratio": 1.4581151832460733,
      "source": "whisper"
    },
    {
      "start": 45.1,
      "end": 47.2,
      "text": "我觉得有的人觉得可能也没法用",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1320807528945635,
      "no_speech_prob": 7.681738578568797e-11,
      "compression_ratio": 1.4581151832460733,
      "source": "whisper"
    },
    {
      "start": 47.2,
      "end": 48.12,
      "text": "MacBook Pro呢",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1320807528945635,
      "no_speech_prob": 7.681738578568797e-11,
      "compression_ratio": 1.4581151832460733,
      "source": "whisper"
    },
    {
      "start": 48.12,
      "end": 49.1,
      "text": "我的一个感觉就是",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1320807528945635,
      "no_speech_prob": 7.681738578568797e-11,
      "compression_ratio": 1.4581151832460733,
      "source": "whisper"
    },
    {
      "start": 49.1,
      "end": 51.56,
      "text": "它的这个噪音控制是确实惊艳",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1320807528945635,
      "no_speech_prob": 7.681738578568797e-11,
      "compression_ratio": 1.4581151832460733,
      "source": "whisper"
    },
    {
      "start": 51.56,
      "end": 53.28,
      "text": "第一部分是最常见的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1320807528945635,
      "no_speech_prob": 7.681738578568797e-11,
      "compression_ratio": 1.4581151832460733,
      "source": "whisper"
    },
    {
      "start": 53.28,
      "end": 54.92,
      "text": "LLM的文字处理能力",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1320807528945635,
      "no_speech_prob": 7.681738578568797e-11,
      "compression_ratio": 1.4581151832460733,
      "source": "whisper"
    },
    {
      "start": 54.92,
      "end": 58.54,
      "text": "那么本次使用的模型具体的有四个",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1320807528945635,
      "no_speech_prob": 7.681738578568797e-11,
      "compression_ratio": 1.4581151832460733,
      "source": "whisper"
    },
    {
      "start": 58.54,
      "end": 62.74,
      "text": "两个GAM4和两个Queen 3.6",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13539742134712837,
      "no_speech_prob": 2.5105887854759423e-11,
      "compression_ratio": 1.2745762711864406,
      "source": "whisper"
    },
    {
      "start": 62.74,
      "end": 65.18,
      "text": "那么它的架构和相关的数据都在这",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13539742134712837,
      "no_speech_prob": 2.5105887854759423e-11,
      "compression_ratio": 1.2745762711864406,
      "source": "whisper"
    },
    {
      "start": 65.18,
      "end": 69.52,
      "text": "发现就是他们的推理速度还是非常可用的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13539742134712837,
      "no_speech_prob": 2.5105887854759423e-11,
      "compression_ratio": 1.2745762711864406,
      "source": "whisper"
    },
    {
      "start": 69.52,
      "end": 74.3,
      "text": "尤其是GAM4的E4B模型和这个千万三的35B MOE模型",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13539742134712837,
      "no_speech_prob": 2.5105887854759423e-11,
      "compression_ratio": 1.2745762711864406,
      "source": "whisper"
    },
    {
      "start": 74.3,
      "end": 78.08,
      "text": "那么虽然这个E4B是一个Dance",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13539742134712837,
      "no_speech_prob": 2.5105887854759423e-11,
      "compression_ratio": 1.2745762711864406,
      "source": "whisper"
    },
    {
      "start": 78.08,
      "end": 79.34,
      "text": "也就是重蜜结构的模型",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13539742134712837,
      "no_speech_prob": 2.5105887854759423e-11,
      "compression_ratio": 1.2745762711864406,
      "source": "whisper"
    },
    {
      "start": 79.34,
      "end": 83.8,
      "text": "但是呢因为它的每次激活是大概有4B的这个参数",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13539742134712837,
      "no_speech_prob": 2.5105887854759423e-11,
      "compression_ratio": 1.2745762711864406,
      "source": "whisper"
    },
    {
      "start": 83.8,
      "end": 85.78,
      "text": "所以它的推理速度也是相当快",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13539742134712837,
      "no_speech_prob": 2.5105887854759423e-11,
      "compression_ratio": 1.2745762711864406,
      "source": "whisper"
    },
    {
      "start": 85.78,
      "end": 87.48,
      "text": "我之前使用的经验",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13539742134712837,
      "no_speech_prob": 2.5105887854759423e-11,
      "compression_ratio": 1.2745762711864406,
      "source": "whisper"
    },
    {
      "start": 87.48,
      "end": 90.52,
      "text": "我发现在具体使用上",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14710636138916017,
      "no_speech_prob": 1.7561242390828902e-11,
      "compression_ratio": 1.3729903536977492,
      "source": "whisper"
    },
    {
      "start": 90.52,
      "end": 92.54,
      "text": "GAMMA模型它的图形的夜能力",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14710636138916017,
      "no_speech_prob": 1.7561242390828902e-11,
      "compression_ratio": 1.3729903536977492,
      "source": "whisper"
    },
    {
      "start": 92.54,
      "end": 94.32,
      "text": "比千万3.6是高一些",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14710636138916017,
      "no_speech_prob": 1.7561242390828902e-11,
      "compression_ratio": 1.3729903536977492,
      "source": "whisper"
    },
    {
      "start": 94.32,
      "end": 97.18,
      "text": "同时我还测量了所谓的月域能力",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14710636138916017,
      "no_speech_prob": 1.7561242390828902e-11,
      "compression_ratio": 1.3729903536977492,
      "source": "whisper"
    },
    {
      "start": 97.18,
      "end": 98.64,
      "text": "也就是BrokerJag的能力",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14710636138916017,
      "no_speech_prob": 1.7561242390828902e-11,
      "compression_ratio": 1.3729903536977492,
      "source": "whisper"
    },
    {
      "start": 98.64,
      "end": 100.64,
      "text": "BrokerJag应该就是方便生成一些",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14710636138916017,
      "no_speech_prob": 1.7561242390828902e-11,
      "compression_ratio": 1.3729903536977492,
      "source": "whisper"
    },
    {
      "start": 100.64,
      "end": 104.48,
      "text": "容易造成误判的一些内容",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14710636138916017,
      "no_speech_prob": 1.7561242390828902e-11,
      "compression_ratio": 1.3729903536977492,
      "source": "whisper"
    },
    {
      "start": 104.48,
      "end": 106.34,
      "text": "包括像一些踩在边界上的内容",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14710636138916017,
      "no_speech_prob": 1.7561242390828902e-11,
      "compression_ratio": 1.3729903536977492,
      "source": "whisper"
    },
    {
      "start": 106.34,
      "end": 108.12,
      "text": "那么在线模型可能生成起来困难",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14710636138916017,
      "no_speech_prob": 1.7561242390828902e-11,
      "compression_ratio": 1.3729903536977492,
      "source": "whisper"
    },
    {
      "start": 108.12,
      "end": 111.0,
      "text": "这个本地生成起来相对就容易很多",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14710636138916017,
      "no_speech_prob": 1.7561242390828902e-11,
      "compression_ratio": 1.3729903536977492,
      "source": "whisper"
    },
    {
      "start": 111.0,
      "end": 112.06,
      "text": "在具体的测试题目上",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14710636138916017,
      "no_speech_prob": 1.7561242390828902e-11,
      "compression_ratio": 1.3729903536977492,
      "source": "whisper"
    },
    {
      "start": 112.06,
      "end": 116.14,
      "text": "我是用这个Cloud生成了一些测试题目",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14710636138916017,
      "no_speech_prob": 1.7561242390828902e-11,
      "compression_ratio": 1.3729903536977492,
      "source": "whisper"
    },
    {
      "start": 116.14,
      "end": 117.92,
      "text": "然后让他帮着完全全自动测试",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08094494505078381,
      "no_speech_prob": 2.7362589624657652e-11,
      "compression_ratio": 1.4378698224852071,
      "source": "whisper"
    },
    {
      "start": 117.92,
      "end": 118.48,
      "text": "内容特别多",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08094494505078381,
      "no_speech_prob": 2.7362589624657652e-11,
      "compression_ratio": 1.4378698224852071,
      "source": "whisper"
    },
    {
      "start": 118.48,
      "end": 121.02,
      "text": "我就不一个一个展示所有的这个答案了",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08094494505078381,
      "no_speech_prob": 2.7362589624657652e-11,
      "compression_ratio": 1.4378698224852071,
      "source": "whisper"
    },
    {
      "start": 121.02,
      "end": 123.54,
      "text": "然后这是一些具体的数据",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08094494505078381,
      "no_speech_prob": 2.7362589624657652e-11,
      "compression_ratio": 1.4378698224852071,
      "source": "whisper"
    },
    {
      "start": 123.54,
      "end": 125.14,
      "text": "总体给我的感觉是这样的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08094494505078381,
      "no_speech_prob": 2.7362589624657652e-11,
      "compression_ratio": 1.4378698224852071,
      "source": "whisper"
    },
    {
      "start": 125.14,
      "end": 127.98,
      "text": "它的题目都能比较好的完成预定的效果",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08094494505078381,
      "no_speech_prob": 2.7362589624657652e-11,
      "compression_ratio": 1.4378698224852071,
      "source": "whisper"
    },
    {
      "start": 127.98,
      "end": 129.62,
      "text": "比如说像有分苹果的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08094494505078381,
      "no_speech_prob": 2.7362589624657652e-11,
      "compression_ratio": 1.4378698224852071,
      "source": "whisper"
    },
    {
      "start": 129.62,
      "end": 131.54,
      "text": "然后分几个的简单的这种盘数",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08094494505078381,
      "no_speech_prob": 2.7362589624657652e-11,
      "compression_ratio": 1.4378698224852071,
      "source": "whisper"
    },
    {
      "start": 131.54,
      "end": 133.38,
      "text": "包括求函数的这种数学题",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08094494505078381,
      "no_speech_prob": 2.7362589624657652e-11,
      "compression_ratio": 1.4378698224852071,
      "source": "whisper"
    },
    {
      "start": 133.38,
      "end": 135.66,
      "text": "或写代码的这种效果",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08094494505078381,
      "no_speech_prob": 2.7362589624657652e-11,
      "compression_ratio": 1.4378698224852071,
      "source": "whisper"
    },
    {
      "start": 135.66,
      "end": 136.52,
      "text": "或修复bug",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08094494505078381,
      "no_speech_prob": 2.7362589624657652e-11,
      "compression_ratio": 1.4378698224852071,
      "source": "whisper"
    },
    {
      "start": 136.52,
      "end": 138.88,
      "text": "包括判断说真假话推理过程",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08094494505078381,
      "no_speech_prob": 2.7362589624657652e-11,
      "compression_ratio": 1.4378698224852071,
      "source": "whisper"
    },
    {
      "start": 138.88,
      "end": 139.84,
      "text": "或解释一些话",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08094494505078381,
      "no_speech_prob": 2.7362589624657652e-11,
      "compression_ratio": 1.4378698224852071,
      "source": "whisper"
    },
    {
      "start": 139.84,
      "end": 142.8,
      "text": "然后用看看具体的执行的能力等等",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08094494505078381,
      "no_speech_prob": 2.7362589624657652e-11,
      "compression_ratio": 1.4378698224852071,
      "source": "whisper"
    },
    {
      "start": 142.8,
      "end": 144.3,
      "text": "就后续的这些答案呢",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08094494505078381,
      "no_speech_prob": 2.7362589624657652e-11,
      "compression_ratio": 1.4378698224852071,
      "source": "whisper"
    },
    {
      "start": 144.3,
      "end": 146.46,
      "text": "我后期将会逐渐整理",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12282626229043332,
      "no_speech_prob": 1.655908049569277e-11,
      "compression_ratio": 1.367132867132867,
      "source": "whisper"
    },
    {
      "start": 146.46,
      "end": 151.46,
      "text": "也去把它放到我这个模型测试的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12282626229043332,
      "no_speech_prob": 1.655908049569277e-11,
      "compression_ratio": 1.367132867132867,
      "source": "whisper"
    },
    {
      "start": 151.46,
      "end": 153.06,
      "text": "这个页面的后边供您参考",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12282626229043332,
      "no_speech_prob": 1.655908049569277e-11,
      "compression_ratio": 1.367132867132867,
      "source": "whisper"
    },
    {
      "start": 153.06,
      "end": 155.54,
      "text": "第二个我测试的是TTS语音",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12282626229043332,
      "no_speech_prob": 1.655908049569277e-11,
      "compression_ratio": 1.367132867132867,
      "source": "whisper"
    },
    {
      "start": 155.54,
      "end": 156.58,
      "text": "语音合成能力",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12282626229043332,
      "no_speech_prob": 1.655908049569277e-11,
      "compression_ratio": 1.367132867132867,
      "source": "whisper"
    },
    {
      "start": 156.58,
      "end": 157.3,
      "text": "这里边呢",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12282626229043332,
      "no_speech_prob": 1.655908049569277e-11,
      "compression_ratio": 1.367132867132867,
      "source": "whisper"
    },
    {
      "start": 157.3,
      "end": 159.58,
      "text": "我目前搜索的资料是这样的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12282626229043332,
      "no_speech_prob": 1.655908049569277e-11,
      "compression_ratio": 1.367132867132867,
      "source": "whisper"
    },
    {
      "start": 159.58,
      "end": 160.5,
      "text": "它有两种架构",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12282626229043332,
      "no_speech_prob": 1.655908049569277e-11,
      "compression_ratio": 1.367132867132867,
      "source": "whisper"
    },
    {
      "start": 160.5,
      "end": 162.78,
      "text": "一个是原声支持MLX",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12282626229043332,
      "no_speech_prob": 1.655908049569277e-11,
      "compression_ratio": 1.367132867132867,
      "source": "whisper"
    },
    {
      "start": 162.78,
      "end": 165.02,
      "text": "这个METAL加速功能的这个架构",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12282626229043332,
      "no_speech_prob": 1.655908049569277e-11,
      "compression_ratio": 1.367132867132867,
      "source": "whisper"
    },
    {
      "start": 165.02,
      "end": 166.68,
      "text": "那么它的速度稍微快一些",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12282626229043332,
      "no_speech_prob": 1.655908049569277e-11,
      "compression_ratio": 1.367132867132867,
      "source": "whisper"
    },
    {
      "start": 166.68,
      "end": 170.46,
      "text": "还有一种是直接用原声这个格式的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12282626229043332,
      "no_speech_prob": 1.655908049569277e-11,
      "compression_ratio": 1.367132867132867,
      "source": "whisper"
    },
    {
      "start": 170.46,
      "end": 174.0,
      "text": "不是支持这个METAL加速的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12282626229043332,
      "no_speech_prob": 1.655908049569277e-11,
      "compression_ratio": 1.367132867132867,
      "source": "whisper"
    },
    {
      "start": 174.0,
      "end": 175.48,
      "text": "那么它俩有一个区别",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11443016745827415,
      "no_speech_prob": 2.6999137300864895e-11,
      "compression_ratio": 1.4027777777777777,
      "source": "whisper"
    },
    {
      "start": 175.48,
      "end": 176.54,
      "text": "原声支持加速的呢",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11443016745827415,
      "no_speech_prob": 2.6999137300864895e-11,
      "compression_ratio": 1.4027777777777777,
      "source": "whisper"
    },
    {
      "start": 176.54,
      "end": 179.04,
      "text": "它在这个进行声音克隆的时候呢",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11443016745827415,
      "no_speech_prob": 2.6999137300864895e-11,
      "compression_ratio": 1.4027777777777777,
      "source": "whisper"
    },
    {
      "start": 179.04,
      "end": 179.9,
      "text": "它是没有那个功能的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11443016745827415,
      "no_speech_prob": 2.6999137300864895e-11,
      "compression_ratio": 1.4027777777777777,
      "source": "whisper"
    },
    {
      "start": 179.9,
      "end": 181.62,
      "text": "只能支持文字描述声音",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11443016745827415,
      "no_speech_prob": 2.6999137300864895e-11,
      "compression_ratio": 1.4027777777777777,
      "source": "whisper"
    },
    {
      "start": 181.62,
      "end": 182.48,
      "text": "那换句话说就是",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11443016745827415,
      "no_speech_prob": 2.6999137300864895e-11,
      "compression_ratio": 1.4027777777777777,
      "source": "whisper"
    },
    {
      "start": 182.48,
      "end": 185.56,
      "text": "你用一段话去让它生成一个你想要的声音",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11443016745827415,
      "no_speech_prob": 2.6999137300864895e-11,
      "compression_ratio": 1.4027777777777777,
      "source": "whisper"
    },
    {
      "start": 185.56,
      "end": 187.14,
      "text": "它根据了你描述的话去生成",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11443016745827415,
      "no_speech_prob": 2.6999137300864895e-11,
      "compression_ratio": 1.4027777777777777,
      "source": "whisper"
    },
    {
      "start": 187.14,
      "end": 189.32,
      "text": "然后速度还是比较快的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11443016745827415,
      "no_speech_prob": 2.6999137300864895e-11,
      "compression_ratio": 1.4027777777777777,
      "source": "whisper"
    },
    {
      "start": 189.32,
      "end": 191.9,
      "text": "它因为没有真正的实现具体声音克隆",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11443016745827415,
      "no_speech_prob": 2.6999137300864895e-11,
      "compression_ratio": 1.4027777777777777,
      "source": "whisper"
    },
    {
      "start": 191.9,
      "end": 195.26,
      "text": "所以只能用另一个叫千万三的Base模型",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11443016745827415,
      "no_speech_prob": 2.6999137300864895e-11,
      "compression_ratio": 1.4027777777777777,
      "source": "whisper"
    },
    {
      "start": 195.26,
      "end": 196.7,
      "text": "这个模型是原版的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11443016745827415,
      "no_speech_prob": 2.6999137300864895e-11,
      "compression_ratio": 1.4027777777777777,
      "source": "whisper"
    },
    {
      "start": 196.7,
      "end": 198.24,
      "text": "它的速度稍微慢一些",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11443016745827415,
      "no_speech_prob": 2.6999137300864895e-11,
      "compression_ratio": 1.4027777777777777,
      "source": "whisper"
    },
    {
      "start": 198.24,
      "end": 199.76,
      "text": "格式呢不是最快的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11443016745827415,
      "no_speech_prob": 2.6999137300864895e-11,
      "compression_ratio": 1.4027777777777777,
      "source": "whisper"
    },
    {
      "start": 199.76,
      "end": 200.66,
      "text": "有意思发现就是",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11443016745827415,
      "no_speech_prob": 2.6999137300864895e-11,
      "compression_ratio": 1.4027777777777777,
      "source": "whisper"
    },
    {
      "start": 200.66,
      "end": 202.44,
      "text": "直接用Voice Design",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11443016745827415,
      "no_speech_prob": 2.6999137300864895e-11,
      "compression_ratio": 1.4027777777777777,
      "source": "whisper"
    },
    {
      "start": 202.44,
      "end": 204.58,
      "text": "也就是描述生成这个音频功能呢",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13002680695575217,
      "no_speech_prob": 2.7722015655262666e-11,
      "compression_ratio": 1.4587155963302751,
      "source": "whisper"
    },
    {
      "start": 204.58,
      "end": 206.64,
      "text": "它的速度反而还比这个声音可能要慢一些",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13002680695575217,
      "no_speech_prob": 2.7722015655262666e-11,
      "compression_ratio": 1.4587155963302751,
      "source": "whisper"
    },
    {
      "start": 206.64,
      "end": 208.02,
      "text": "不知道这是因为",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13002680695575217,
      "no_speech_prob": 2.7722015655262666e-11,
      "compression_ratio": 1.4587155963302751,
      "source": "whisper"
    },
    {
      "start": 208.02,
      "end": 209.12,
      "text": "为什么",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13002680695575217,
      "no_speech_prob": 2.7722015655262666e-11,
      "compression_ratio": 1.4587155963302751,
      "source": "whisper"
    },
    {
      "start": 209.12,
      "end": 210.44,
      "text": "可能是因为处理架构上的音",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13002680695575217,
      "no_speech_prob": 2.7722015655262666e-11,
      "compression_ratio": 1.4587155963302751,
      "source": "whisper"
    },
    {
      "start": 210.44,
      "end": 211.66,
      "text": "需要注意的是它有一个bug",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13002680695575217,
      "no_speech_prob": 2.7722015655262666e-11,
      "compression_ratio": 1.4587155963302751,
      "source": "whisper"
    },
    {
      "start": 211.66,
      "end": 214.92,
      "text": "就是截取的那个具体参考的参数",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13002680695575217,
      "no_speech_prob": 2.7722015655262666e-11,
      "compression_ratio": 1.4587155963302751,
      "source": "whisper"
    },
    {
      "start": 214.92,
      "end": 217.36,
      "text": "所你截取的音频是需要匹配",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13002680695575217,
      "no_speech_prob": 2.7722015655262666e-11,
      "compression_ratio": 1.4587155963302751,
      "source": "whisper"
    },
    {
      "start": 217.36,
      "end": 219.26,
      "text": "另外一个五秒即可",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13002680695575217,
      "no_speech_prob": 2.7722015655262666e-11,
      "compression_ratio": 1.4587155963302751,
      "source": "whisper"
    },
    {
      "start": 219.26,
      "end": 221.14,
      "text": "如果这个内容特别多",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13002680695575217,
      "no_speech_prob": 2.7722015655262666e-11,
      "compression_ratio": 1.4587155963302751,
      "source": "whisper"
    },
    {
      "start": 221.14,
      "end": 222.3,
      "text": "或者它两者比匹配的话",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13002680695575217,
      "no_speech_prob": 2.7722015655262666e-11,
      "compression_ratio": 1.4587155963302751,
      "source": "whisper"
    },
    {
      "start": 222.3,
      "end": 223.24,
      "text": "会生成一些乱码",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13002680695575217,
      "no_speech_prob": 2.7722015655262666e-11,
      "compression_ratio": 1.4587155963302751,
      "source": "whisper"
    },
    {
      "start": 223.24,
      "end": 224.06,
      "text": "真人的录音",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13002680695575217,
      "no_speech_prob": 2.7722015655262666e-11,
      "compression_ratio": 1.4587155963302751,
      "source": "whisper"
    },
    {
      "start": 224.06,
      "end": 225.06,
      "text": "我们听一听这个录音",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13002680695575217,
      "no_speech_prob": 2.7722015655262666e-11,
      "compression_ratio": 1.4587155963302751,
      "source": "whisper"
    },
    {
      "start": 225.06,
      "end": 226.3,
      "text": "叫做装算赤",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13002680695575217,
      "no_speech_prob": 2.7722015655262666e-11,
      "compression_ratio": 1.4587155963302751,
      "source": "whisper"
    },
    {
      "start": 226.3,
      "end": 227.16,
      "text": "副算黑",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13002680695575217,
      "no_speech_prob": 2.7722015655262666e-11,
      "compression_ratio": 1.4587155963302751,
      "source": "whisper"
    },
    {
      "start": 227.16,
      "end": 229.08,
      "text": "赤就是红色",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13002680695575217,
      "no_speech_prob": 2.7722015655262666e-11,
      "compression_ratio": 1.4587155963302751,
      "source": "whisper"
    },
    {
      "start": 229.08,
      "end": 230.4,
      "text": "我们这次听一下",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13002680695575217,
      "no_speech_prob": 2.7722015655262666e-11,
      "compression_ratio": 1.4587155963302751,
      "source": "whisper"
    },
    {
      "start": 230.4,
      "end": 232.9,
      "text": "它克隆完毕之后的声音的感觉",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17548999786376954,
      "no_speech_prob": 3.0708723758321455e-11,
      "compression_ratio": 1.1778656126482214,
      "source": "whisper"
    },
    {
      "start": 232.9,
      "end": 243.94,
      "text": "大家可以听到",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17548999786376954,
      "no_speech_prob": 3.0708723758321455e-11,
      "compression_ratio": 1.1778656126482214,
      "source": "whisper"
    },
    {
      "start": 243.94,
      "end": 245.88,
      "text": "这个克隆的声音简直是一家论真",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17548999786376954,
      "no_speech_prob": 3.0708723758321455e-11,
      "compression_ratio": 1.1778656126482214,
      "source": "whisper"
    },
    {
      "start": 245.88,
      "end": 247.62,
      "text": "几乎听不出任何的区别",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17548999786376954,
      "no_speech_prob": 3.0708723758321455e-11,
      "compression_ratio": 1.1778656126482214,
      "source": "whisper"
    },
    {
      "start": 247.62,
      "end": 249.06,
      "text": "TTLS语音合成功能上",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17548999786376954,
      "no_speech_prob": 3.0708723758321455e-11,
      "compression_ratio": 1.1778656126482214,
      "source": "whisper"
    },
    {
      "start": 249.06,
      "end": 251.02,
      "text": "我觉得还是相当强大",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17548999786376954,
      "no_speech_prob": 3.0708723758321455e-11,
      "compression_ratio": 1.1778656126482214,
      "source": "whisper"
    },
    {
      "start": 251.02,
      "end": 251.48,
      "text": "第三点",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17548999786376954,
      "no_speech_prob": 3.0708723758321455e-11,
      "compression_ratio": 1.1778656126482214,
      "source": "whisper"
    },
    {
      "start": 251.48,
      "end": 252.54,
      "text": "看图片生成能力",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17548999786376954,
      "no_speech_prob": 3.0708723758321455e-11,
      "compression_ratio": 1.1778656126482214,
      "source": "whisper"
    },
    {
      "start": 252.54,
      "end": 254.84,
      "text": "我这里选用的是质量非常高的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17548999786376954,
      "no_speech_prob": 3.0708723758321455e-11,
      "compression_ratio": 1.1778656126482214,
      "source": "whisper"
    },
    {
      "start": 254.84,
      "end": 255.12,
      "text": "这个",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17548999786376954,
      "no_speech_prob": 3.0708723758321455e-11,
      "compression_ratio": 1.1778656126482214,
      "source": "whisper"
    },
    {
      "start": 255.12,
      "end": 258.32,
      "text": "Fluzy1的DVE的模型",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17548999786376954,
      "no_speech_prob": 3.0708723758321455e-11,
      "compression_ratio": 1.1778656126482214,
      "source": "whisper"
    },
    {
      "start": 258.32,
      "end": 259.48,
      "text": "后侧的是两方面",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17548999786376954,
      "no_speech_prob": 3.0708723758321455e-11,
      "compression_ratio": 1.1778656126482214,
      "source": "whisper"
    },
    {
      "start": 259.48,
      "end": 261.2,
      "text": "一个是直接生成图片",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10126977633935799,
      "no_speech_prob": 2.8603525797921137e-11,
      "compression_ratio": 1.5147058823529411,
      "source": "whisper"
    },
    {
      "start": 261.2,
      "end": 262.6,
      "text": "另一个是使用Laura",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10126977633935799,
      "no_speech_prob": 2.8603525797921137e-11,
      "compression_ratio": 1.5147058823529411,
      "source": "whisper"
    },
    {
      "start": 262.6,
      "end": 264.46,
      "text": "那么具体的生成时间在这里",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10126977633935799,
      "no_speech_prob": 2.8603525797921137e-11,
      "compression_ratio": 1.5147058823529411,
      "source": "whisper"
    },
    {
      "start": 264.46,
      "end": 266.08,
      "text": "风景这个图片生成的时候呢",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10126977633935799,
      "no_speech_prob": 2.8603525797921137e-11,
      "compression_ratio": 1.5147058823529411,
      "source": "whisper"
    },
    {
      "start": 266.08,
      "end": 268.44,
      "text": "它因为需要进行这个冷启动",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10126977633935799,
      "no_speech_prob": 2.8603525797921137e-11,
      "compression_ratio": 1.5147058823529411,
      "source": "whisper"
    },
    {
      "start": 268.44,
      "end": 269.5,
      "text": "然后注入参数",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10126977633935799,
      "no_speech_prob": 2.8603525797921137e-11,
      "compression_ratio": 1.5147058823529411,
      "source": "whisper"
    },
    {
      "start": 269.5,
      "end": 270.86,
      "text": "所以它的生成速度是最慢的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10126977633935799,
      "no_speech_prob": 2.8603525797921137e-11,
      "compression_ratio": 1.5147058823529411,
      "source": "whisper"
    },
    {
      "start": 270.86,
      "end": 272.8,
      "text": "然后接下来速度就相对来说稳定很多",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10126977633935799,
      "no_speech_prob": 2.8603525797921137e-11,
      "compression_ratio": 1.5147058823529411,
      "source": "whisper"
    },
    {
      "start": 272.8,
      "end": 274.78,
      "text": "大概生成一幅像1024x1024",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10126977633935799,
      "no_speech_prob": 2.8603525797921137e-11,
      "compression_ratio": 1.5147058823529411,
      "source": "whisper"
    },
    {
      "start": 274.78,
      "end": 277.06,
      "text": "也就是它的分辨率相对较高的情况下",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10126977633935799,
      "no_speech_prob": 2.8603525797921137e-11,
      "compression_ratio": 1.5147058823529411,
      "source": "whisper"
    },
    {
      "start": 277.06,
      "end": 279.66,
      "text": "每一幅的图片大概是用到的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10126977633935799,
      "no_speech_prob": 2.8603525797921137e-11,
      "compression_ratio": 1.5147058823529411,
      "source": "whisper"
    },
    {
      "start": 279.66,
      "end": 280.92,
      "text": "71分到11分",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10126977633935799,
      "no_speech_prob": 2.8603525797921137e-11,
      "compression_ratio": 1.5147058823529411,
      "source": "whisper"
    },
    {
      "start": 280.92,
      "end": 282.74,
      "text": "直接用Laura进行稳定",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10126977633935799,
      "no_speech_prob": 2.8603525797921137e-11,
      "compression_ratio": 1.5147058823529411,
      "source": "whisper"
    },
    {
      "start": 282.74,
      "end": 284.64,
      "text": "这个真人图片克隆的情况下",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10126977633935799,
      "no_speech_prob": 2.8603525797921137e-11,
      "compression_ratio": 1.5147058823529411,
      "source": "whisper"
    },
    {
      "start": 284.64,
      "end": 286.3,
      "text": "也是大概这个相同的速度",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10126977633935799,
      "no_speech_prob": 2.8603525797921137e-11,
      "compression_ratio": 1.5147058823529411,
      "source": "whisper"
    },
    {
      "start": 286.3,
      "end": 286.84,
      "text": "看一下结果",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10126977633935799,
      "no_speech_prob": 2.8603525797921137e-11,
      "compression_ratio": 1.5147058823529411,
      "source": "whisper"
    },
    {
      "start": 286.84,
      "end": 287.68,
      "text": "第一个是风景",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.10126977633935799,
      "no_speech_prob": 2.8603525797921137e-11,
      "compression_ratio": 1.5147058823529411,
      "source": "whisper"
    },
    {
      "start": 287.68,
      "end": 290.74,
      "text": "我们看到这些细节给人的感觉还是非常不错的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08487404326689409,
      "no_speech_prob": 2.7628200074958364e-11,
      "compression_ratio": 1.4415954415954415,
      "source": "whisper"
    },
    {
      "start": 290.74,
      "end": 291.22,
      "text": "效果很好",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08487404326689409,
      "no_speech_prob": 2.7628200074958364e-11,
      "compression_ratio": 1.4415954415954415,
      "source": "whisper"
    },
    {
      "start": 291.22,
      "end": 292.3,
      "text": "赛博朋克的业绩",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08487404326689409,
      "no_speech_prob": 2.7628200074958364e-11,
      "compression_ratio": 1.4415954415954415,
      "source": "whisper"
    },
    {
      "start": 292.3,
      "end": 293.22,
      "text": "专业商务人像",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08487404326689409,
      "no_speech_prob": 2.7628200074958364e-11,
      "compression_ratio": 1.4415954415954415,
      "source": "whisper"
    },
    {
      "start": 293.22,
      "end": 294.44,
      "text": "比较抽象的液态金属",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08487404326689409,
      "no_speech_prob": 2.7628200074958364e-11,
      "compression_ratio": 1.4415954415954415,
      "source": "whisper"
    },
    {
      "start": 294.44,
      "end": 295.34,
      "text": "产品设计度光影",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08487404326689409,
      "no_speech_prob": 2.7628200074958364e-11,
      "compression_ratio": 1.4415954415954415,
      "source": "whisper"
    },
    {
      "start": 295.34,
      "end": 296.32,
      "text": "还是非常不错的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08487404326689409,
      "no_speech_prob": 2.7628200074958364e-11,
      "compression_ratio": 1.4415954415954415,
      "source": "whisper"
    },
    {
      "start": 296.32,
      "end": 299.52,
      "text": "比较好的执行了我需要的这个要求",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08487404326689409,
      "no_speech_prob": 2.7628200074958364e-11,
      "compression_ratio": 1.4415954415954415,
      "source": "whisper"
    },
    {
      "start": 299.52,
      "end": 301.7,
      "text": "我之前没有任何训练经验",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08487404326689409,
      "no_speech_prob": 2.7628200074958364e-11,
      "compression_ratio": 1.4415954415954415,
      "source": "whisper"
    },
    {
      "start": 301.7,
      "end": 302.38,
      "text": "所以我就自己",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08487404326689409,
      "no_speech_prob": 2.7628200074958364e-11,
      "compression_ratio": 1.4415954415954415,
      "source": "whisper"
    },
    {
      "start": 302.38,
      "end": 305.26,
      "text": "我就自己随意拍了很多照片",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08487404326689409,
      "no_speech_prob": 2.7628200074958364e-11,
      "compression_ratio": 1.4415954415954415,
      "source": "whisper"
    },
    {
      "start": 305.26,
      "end": 306.12,
      "text": "各种角度",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08487404326689409,
      "no_speech_prob": 2.7628200074958364e-11,
      "compression_ratio": 1.4415954415954415,
      "source": "whisper"
    },
    {
      "start": 306.12,
      "end": 307.42,
      "text": "各种的方向",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08487404326689409,
      "no_speech_prob": 2.7628200074958364e-11,
      "compression_ratio": 1.4415954415954415,
      "source": "whisper"
    },
    {
      "start": 307.42,
      "end": 308.22,
      "text": "然后各种场景",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08487404326689409,
      "no_speech_prob": 2.7628200074958364e-11,
      "compression_ratio": 1.4415954415954415,
      "source": "whisper"
    },
    {
      "start": 308.22,
      "end": 309.28,
      "text": "尽量穿不同的衣服",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08487404326689409,
      "no_speech_prob": 2.7628200074958364e-11,
      "compression_ratio": 1.4415954415954415,
      "source": "whisper"
    },
    {
      "start": 309.28,
      "end": 312.14,
      "text": "然后装束尽量去不一样",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08487404326689409,
      "no_speech_prob": 2.7628200074958364e-11,
      "compression_ratio": 1.4415954415954415,
      "source": "whisper"
    },
    {
      "start": 312.14,
      "end": 313.92,
      "text": "然后拍摄了远近各种各样的照片",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08487404326689409,
      "no_speech_prob": 2.7628200074958364e-11,
      "compression_ratio": 1.4415954415954415,
      "source": "whisper"
    },
    {
      "start": 313.92,
      "end": 315.12,
      "text": "数量是60多张",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08487404326689409,
      "no_speech_prob": 2.7628200074958364e-11,
      "compression_ratio": 1.4415954415954415,
      "source": "whisper"
    },
    {
      "start": 315.12,
      "end": 315.84,
      "text": "然后加在一块",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08487404326689409,
      "no_speech_prob": 2.7628200074958364e-11,
      "compression_ratio": 1.4415954415954415,
      "source": "whisper"
    },
    {
      "start": 315.84,
      "end": 316.78,
      "text": "所用的触发词",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08487404326689409,
      "no_speech_prob": 2.7628200074958364e-11,
      "compression_ratio": 1.4415954415954415,
      "source": "whisper"
    },
    {
      "start": 316.78,
      "end": 320.18,
      "text": "这里的触发词就是一个所谓比较生辟的词",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14258343166775173,
      "no_speech_prob": 3.7192221524762203e-11,
      "compression_ratio": 1.4282051282051282,
      "source": "whisper"
    },
    {
      "start": 320.18,
      "end": 321.34,
      "text": "然后帮助触发这个人像",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14258343166775173,
      "no_speech_prob": 3.7192221524762203e-11,
      "compression_ratio": 1.4282051282051282,
      "source": "whisper"
    },
    {
      "start": 321.34,
      "end": 323.38,
      "text": "使用的相关参数是scale等1.0",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14258343166775173,
      "no_speech_prob": 3.7192221524762203e-11,
      "compression_ratio": 1.4282051282051282,
      "source": "whisper"
    },
    {
      "start": 323.38,
      "end": 324.32,
      "text": "是第一张商务头像",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14258343166775173,
      "no_speech_prob": 3.7192221524762203e-11,
      "compression_ratio": 1.4282051282051282,
      "source": "whisper"
    },
    {
      "start": 324.32,
      "end": 325.32,
      "text": "这个可以看起来",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14258343166775173,
      "no_speech_prob": 3.7192221524762203e-11,
      "compression_ratio": 1.4282051282051282,
      "source": "whisper"
    },
    {
      "start": 325.32,
      "end": 326.68,
      "text": "就是样子是不戴眼镜的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14258343166775173,
      "no_speech_prob": 3.7192221524762203e-11,
      "compression_ratio": 1.4282051282051282,
      "source": "whisper"
    },
    {
      "start": 326.68,
      "end": 328.48,
      "text": "是我本人可能有这个送眉的习惯",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14258343166775173,
      "no_speech_prob": 3.7192221524762203e-11,
      "compression_ratio": 1.4282051282051282,
      "source": "whisper"
    },
    {
      "start": 328.48,
      "end": 329.28,
      "text": "它这个地方呢",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14258343166775173,
      "no_speech_prob": 3.7192221524762203e-11,
      "compression_ratio": 1.4282051282051282,
      "source": "whisper"
    },
    {
      "start": 329.28,
      "end": 330.82,
      "text": "就特别明显的捕捉到",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14258343166775173,
      "no_speech_prob": 3.7192221524762203e-11,
      "compression_ratio": 1.4282051282051282,
      "source": "whisper"
    },
    {
      "start": 330.82,
      "end": 332.06,
      "text": "然后剩下两边",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14258343166775173,
      "no_speech_prob": 3.7192221524762203e-11,
      "compression_ratio": 1.4282051282051282,
      "source": "whisper"
    },
    {
      "start": 332.06,
      "end": 336.9,
      "text": "我这个因为现在正在处于一直控制提供的状态下",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14258343166775173,
      "no_speech_prob": 3.7192221524762203e-11,
      "compression_ratio": 1.4282051282051282,
      "source": "whisper"
    },
    {
      "start": 336.9,
      "end": 338.14,
      "text": "所以胖瘦呢有时候不太一定",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14258343166775173,
      "no_speech_prob": 3.7192221524762203e-11,
      "compression_ratio": 1.4282051282051282,
      "source": "whisper"
    },
    {
      "start": 338.14,
      "end": 340.24,
      "text": "像这张照片就生成的稍微胖一点",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14258343166775173,
      "no_speech_prob": 3.7192221524762203e-11,
      "compression_ratio": 1.4282051282051282,
      "source": "whisper"
    },
    {
      "start": 340.24,
      "end": 341.44,
      "text": "然后这张也是",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14258343166775173,
      "no_speech_prob": 3.7192221524762203e-11,
      "compression_ratio": 1.4282051282051282,
      "source": "whisper"
    },
    {
      "start": 341.44,
      "end": 342.02,
      "text": "这是戴眼镜",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14258343166775173,
      "no_speech_prob": 3.7192221524762203e-11,
      "compression_ratio": 1.4282051282051282,
      "source": "whisper"
    },
    {
      "start": 342.02,
      "end": 343.28,
      "text": "大家看到我俩现在的样子",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14258343166775173,
      "no_speech_prob": 3.7192221524762203e-11,
      "compression_ratio": 1.4282051282051282,
      "source": "whisper"
    },
    {
      "start": 343.28,
      "end": 343.82,
      "text": "就先回来的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14258343166775173,
      "no_speech_prob": 3.7192221524762203e-11,
      "compression_ratio": 1.4282051282051282,
      "source": "whisper"
    },
    {
      "start": 343.82,
      "end": 344.66,
      "text": "这个是科技演讲",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14258343166775173,
      "no_speech_prob": 3.7192221524762203e-11,
      "compression_ratio": 1.4282051282051282,
      "source": "whisper"
    },
    {
      "start": 344.66,
      "end": 345.62,
      "text": "然后科技演讲的呢",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 345.62,
      "end": 346.7,
      "text": "这个侧面我发现的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 346.7,
      "end": 348.06,
      "text": "因为可能我给它的样子",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 348.06,
      "end": 349.14,
      "text": "侧面的照片特别多",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 349.14,
      "end": 349.74,
      "text": "所以这个侧面",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 349.74,
      "end": 350.86,
      "text": "我觉得还非常非常像",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 350.86,
      "end": 351.8,
      "text": "然后但是有一个问题",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 351.8,
      "end": 353.48,
      "text": "它出现了比较严重的幻觉",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 353.48,
      "end": 355.24,
      "text": "就是这个手出现了多个",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 355.24,
      "end": 356.02,
      "text": "然后并且胸口",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 356.02,
      "end": 357.34,
      "text": "这有一个我给它的提示词",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 357.34,
      "end": 358.1,
      "text": "就伸透出来了",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 358.1,
      "end": 358.62,
      "text": "HIFEN TAG",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 358.62,
      "end": 359.6,
      "text": "这次生成呢",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 359.6,
      "end": 360.5,
      "text": "我是直接",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 360.5,
      "end": 361.74,
      "text": "不管它生成出什么来",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 361.74,
      "end": 362.68,
      "text": "我全原样展示出来",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 362.68,
      "end": 363.94,
      "text": "所以这是不是经过我选择的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 363.94,
      "end": 365.58,
      "text": "这就是直接一次性的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 365.58,
      "end": 366.16,
      "text": "所以等于",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 366.16,
      "end": 367.86,
      "text": "我觉得成功率相对来说",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 367.86,
      "end": 368.58,
      "text": "还是可用的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 368.58,
      "end": 369.4,
      "text": "非常可用的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 369.4,
      "end": 370.24,
      "text": "可能得分场景",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 370.24,
      "end": 371.48,
      "text": "然后可能稍微抽抽卡",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 371.48,
      "end": 372.44,
      "text": "一系列的大家看到",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15344789293077257,
      "no_speech_prob": 4.591142743759491e-11,
      "compression_ratio": 1.5347043701799485,
      "source": "whisper"
    },
    {
      "start": 374.66,
      "end": 375.3,
      "text": "一个人像的样子",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 375.3,
      "end": 377.7,
      "text": "基本上是大概六成吧",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 377.7,
      "end": 379.26,
      "text": "我把四成这个删除了",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 379.26,
      "end": 380.5,
      "text": "六成之后生成出来的样子",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 380.5,
      "end": 380.74,
      "text": "是这样",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 380.74,
      "end": 381.5,
      "text": "视频生成这个部分",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 381.5,
      "end": 382.42,
      "text": "那么大致是两类",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 382.42,
      "end": 383.22,
      "text": "一类是完成视频",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 383.22,
      "end": 384.82,
      "text": "还有一类是图生视频",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 384.82,
      "end": 385.78,
      "text": "然后它的方式呢",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 385.78,
      "end": 386.54,
      "text": "有两类",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 386.54,
      "end": 387.3,
      "text": "一类是蒸馏模式",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 387.3,
      "end": 387.7,
      "text": "速度快",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 387.7,
      "end": 388.42,
      "text": "还有一类是2C",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 388.42,
      "end": 389.14,
      "text": "是两阶段模式",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 389.14,
      "end": 389.82,
      "text": "速度迈起",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 389.82,
      "end": 390.98,
      "text": "它俩之间的速度相差",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 390.98,
      "end": 391.5,
      "text": "比较明显的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 391.5,
      "end": 392.54,
      "text": "左侧是蒸馏版",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 392.54,
      "end": 393.78,
      "text": "右侧是这个",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 393.78,
      "end": 395.5,
      "text": "所谓更加精细的两阶段版本",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 395.5,
      "end": 397.1,
      "text": "速度相差45倍吧",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 397.1,
      "end": 398.54,
      "text": "两种版本各生成两",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 398.54,
      "end": 399.58,
      "text": "我们先说是第一个",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.20633768717447917,
      "no_speech_prob": 5.2372189918159506e-11,
      "compression_ratio": 1.5638629283489096,
      "source": "whisper"
    },
    {
      "start": 399.58,
      "end": 401.58,
      "text": "城市街道这个是蒸油版生成的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17147644953941232,
      "no_speech_prob": 5.344193837464317e-11,
      "compression_ratio": 1.6232876712328768,
      "source": "whisper"
    },
    {
      "start": 401.58,
      "end": 404.58,
      "text": "明显描述是让它生成城市街道",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17147644953941232,
      "no_speech_prob": 5.344193837464317e-11,
      "compression_ratio": 1.6232876712328768,
      "source": "whisper"
    },
    {
      "start": 404.58,
      "end": 405.9,
      "text": "但是它生成了一个室内景",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17147644953941232,
      "no_speech_prob": 5.344193837464317e-11,
      "compression_ratio": 1.6232876712328768,
      "source": "whisper"
    },
    {
      "start": 405.9,
      "end": 407.58,
      "text": "有车库刚才有白色人物",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17147644953941232,
      "no_speech_prob": 5.344193837464317e-11,
      "compression_ratio": 1.6232876712328768,
      "source": "whisper"
    },
    {
      "start": 407.58,
      "end": 409.58,
      "text": "跟城市街道的明显偏离了",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17147644953941232,
      "no_speech_prob": 5.344193837464317e-11,
      "compression_ratio": 1.6232876712328768,
      "source": "whisper"
    },
    {
      "start": 409.58,
      "end": 410.98,
      "text": "所以理解确实有问题",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17147644953941232,
      "no_speech_prob": 5.344193837464317e-11,
      "compression_ratio": 1.6232876712328768,
      "source": "whisper"
    },
    {
      "start": 410.98,
      "end": 412.58,
      "text": "第二个城市街道呢",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17147644953941232,
      "no_speech_prob": 5.344193837464317e-11,
      "compression_ratio": 1.6232876712328768,
      "source": "whisper"
    },
    {
      "start": 412.58,
      "end": 413.58,
      "text": "两街道版本",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17147644953941232,
      "no_speech_prob": 5.344193837464317e-11,
      "compression_ratio": 1.6232876712328768,
      "source": "whisper"
    },
    {
      "start": 413.58,
      "end": 415.58,
      "text": "确实生成的街道有了提升",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17147644953941232,
      "no_speech_prob": 5.344193837464317e-11,
      "compression_ratio": 1.6232876712328768,
      "source": "whisper"
    },
    {
      "start": 415.58,
      "end": 416.58,
      "text": "然后有人了",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17147644953941232,
      "no_speech_prob": 5.344193837464317e-11,
      "compression_ratio": 1.6232876712328768,
      "source": "whisper"
    },
    {
      "start": 416.58,
      "end": 417.58,
      "text": "然后有人群了",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17147644953941232,
      "no_speech_prob": 5.344193837464317e-11,
      "compression_ratio": 1.6232876712328768,
      "source": "whisper"
    },
    {
      "start": 417.58,
      "end": 418.58,
      "text": "有建筑物了",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17147644953941232,
      "no_speech_prob": 5.344193837464317e-11,
      "compression_ratio": 1.6232876712328768,
      "source": "whisper"
    },
    {
      "start": 418.58,
      "end": 420.58,
      "text": "然后时间也增大了很多",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17147644953941232,
      "no_speech_prob": 5.344193837464317e-11,
      "compression_ratio": 1.6232876712328768,
      "source": "whisper"
    },
    {
      "start": 420.58,
      "end": 422.58,
      "text": "但是这是夜间警官",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17147644953941232,
      "no_speech_prob": 5.344193837464317e-11,
      "compression_ratio": 1.6232876712328768,
      "source": "whisper"
    },
    {
      "start": 422.58,
      "end": 423.58,
      "text": "它稍微理解有点问题",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17147644953941232,
      "no_speech_prob": 5.344193837464317e-11,
      "compression_ratio": 1.6232876712328768,
      "source": "whisper"
    },
    {
      "start": 423.58,
      "end": 424.58,
      "text": "这点是有点缺陷",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17147644953941232,
      "no_speech_prob": 5.344193837464317e-11,
      "compression_ratio": 1.6232876712328768,
      "source": "whisper"
    },
    {
      "start": 424.58,
      "end": 426.58,
      "text": "第二个是科技实验室",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17147644953941232,
      "no_speech_prob": 5.344193837464317e-11,
      "compression_ratio": 1.6232876712328768,
      "source": "whisper"
    },
    {
      "start": 426.58,
      "end": 427.58,
      "text": "在科技实验室里边",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.17147644953941232,
      "no_speech_prob": 5.344193837464317e-11,
      "compression_ratio": 1.6232876712328768,
      "source": "whisper"
    },
    {
      "start": 427.58,
      "end": 430.34,
      "text": "这个声唱令利母女人剩下完全的要素",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.31980022430419924,
      "no_speech_prob": 2.9034535192762334e-11,
      "compression_ratio": 1.0,
      "source": "whisper"
    },
    {
      "start": 430.34,
      "end": 431.86,
      "text": "基本上没有体现出来",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.31980022430419924,
      "no_speech_prob": 2.9034535192762334e-11,
      "compression_ratio": 1.0,
      "source": "whisper"
    },
    {
      "start": 431.86,
      "end": 434.74,
      "text": "像手啊键盘啊灯光啊实验室这种感觉啊",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.31980022430419924,
      "no_speech_prob": 2.9034535192762334e-11,
      "compression_ratio": 1.0,
      "source": "whisper"
    },
    {
      "start": 434.805,
      "end": 440.805,
      "text": "後兩階的版本呢,生成了一個這樣的視頻。一個類似於科技人員把它腳化。有點像保安或者快遞啊,外國。",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.28334173864247847,
      "no_speech_prob": 1.08981721216872e-11,
      "compression_ratio": 1.4905149051490514,
      "source": "whisper"
    },
    {
      "start": 440.805,
      "end": 445.805,
      "text": "然後它生成了這些像手臂啊、屏幕這種樣子。接近了這個描述的樣子。",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.28334173864247847,
      "no_speech_prob": 1.08981721216872e-11,
      "compression_ratio": 1.4905149051490514,
      "source": "whisper"
    },
    {
      "start": 445.805,
      "end": 447.805,
      "text": "但是核心动作包括这种感觉是没有的",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10022082858615451,
      "no_speech_prob": 7.715701688670862e-11,
      "compression_ratio": 1.434782608695652,
      "source": "whisper"
    },
    {
      "start": 447.805,
      "end": 448.805,
      "text": "我也发现了",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10022082858615451,
      "no_speech_prob": 7.715701688670862e-11,
      "compression_ratio": 1.434782608695652,
      "source": "whisper"
    },
    {
      "start": 448.805,
      "end": 451.805,
      "text": "好像fts模型对于生成纯风景",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10022082858615451,
      "no_speech_prob": 7.715701688670862e-11,
      "compression_ratio": 1.434782608695652,
      "source": "whisper"
    },
    {
      "start": 451.805,
      "end": 454.805,
      "text": "或者纯那种建筑物之类的那种模式",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10022082858615451,
      "no_speech_prob": 7.715701688670862e-11,
      "compression_ratio": 1.434782608695652,
      "source": "whisper"
    },
    {
      "start": 454.805,
      "end": 455.805,
      "text": "他经常愿意加人进去",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10022082858615451,
      "no_speech_prob": 7.715701688670862e-11,
      "compression_ratio": 1.434782608695652,
      "source": "whisper"
    },
    {
      "start": 455.805,
      "end": 456.805,
      "text": "就每一个都要有点人",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10022082858615451,
      "no_speech_prob": 7.715701688670862e-11,
      "compression_ratio": 1.434782608695652,
      "source": "whisper"
    },
    {
      "start": 456.805,
      "end": 457.805,
      "text": "可能比较富有人性",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10022082858615451,
      "no_speech_prob": 7.715701688670862e-11,
      "compression_ratio": 1.434782608695652,
      "source": "whisper"
    },
    {
      "start": 457.805,
      "end": 458.805,
      "text": "这是他的问题",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10022082858615451,
      "no_speech_prob": 7.715701688670862e-11,
      "compression_ratio": 1.434782608695652,
      "source": "whisper"
    },
    {
      "start": 458.805,
      "end": 459.805,
      "text": "为了去改变这个问题",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10022082858615451,
      "no_speech_prob": 7.715701688670862e-11,
      "compression_ratio": 1.434782608695652,
      "source": "whisper"
    },
    {
      "start": 459.805,
      "end": 461.805,
      "text": "我发现的办法就是用无声视频的方式",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10022082858615451,
      "no_speech_prob": 7.715701688670862e-11,
      "compression_ratio": 1.434782608695652,
      "source": "whisper"
    },
    {
      "start": 461.805,
      "end": 462.805,
      "text": "给他一些图",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10022082858615451,
      "no_speech_prob": 7.715701688670862e-11,
      "compression_ratio": 1.434782608695652,
      "source": "whisper"
    },
    {
      "start": 462.805,
      "end": 463.805,
      "text": "然后让他去发挥",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10022082858615451,
      "no_speech_prob": 7.715701688670862e-11,
      "compression_ratio": 1.434782608695652,
      "source": "whisper"
    },
    {
      "start": 463.805,
      "end": 464.805,
      "text": "那样的话效果会好很多",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10022082858615451,
      "no_speech_prob": 7.715701688670862e-11,
      "compression_ratio": 1.434782608695652,
      "source": "whisper"
    },
    {
      "start": 464.805,
      "end": 465.805,
      "text": "无声视频",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10022082858615451,
      "no_speech_prob": 7.715701688670862e-11,
      "compression_ratio": 1.434782608695652,
      "source": "whisper"
    },
    {
      "start": 465.805,
      "end": 466.805,
      "text": "然后让图片说话",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10022082858615451,
      "no_speech_prob": 7.715701688670862e-11,
      "compression_ratio": 1.434782608695652,
      "source": "whisper"
    },
    {
      "start": 466.805,
      "end": 468.805,
      "text": "找一张随机图库去测试",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10022082858615451,
      "no_speech_prob": 7.715701688670862e-11,
      "compression_ratio": 1.434782608695652,
      "source": "whisper"
    },
    {
      "start": 468.805,
      "end": 469.805,
      "text": "可是因为他是自动测试",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10022082858615451,
      "no_speech_prob": 7.715701688670862e-11,
      "compression_ratio": 1.434782608695652,
      "source": "whisper"
    },
    {
      "start": 469.805,
      "end": 470.805,
      "text": "我们有干预",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10022082858615451,
      "no_speech_prob": 7.715701688670862e-11,
      "compression_ratio": 1.434782608695652,
      "source": "whisper"
    },
    {
      "start": 470.805,
      "end": 471.805,
      "text": "他测试的时候找他",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10022082858615451,
      "no_speech_prob": 7.715701688670862e-11,
      "compression_ratio": 1.434782608695652,
      "source": "whisper"
    },
    {
      "start": 471.805,
      "end": 472.805,
      "text": "本来说想说一个人让他说话",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10022082858615451,
      "no_speech_prob": 7.715701688670862e-11,
      "compression_ratio": 1.434782608695652,
      "source": "whisper"
    },
    {
      "start": 472.805,
      "end": 473.645,
      "text": "说一个人让他说话",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.24606865776909723,
      "no_speech_prob": 4.710152406994794e-11,
      "compression_ratio": 1.4747191011235956,
      "source": "whisper"
    },
    {
      "start": 473.645,
      "end": 474.765,
      "text": "结果找成了一只狗",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.24606865776909723,
      "no_speech_prob": 4.710152406994794e-11,
      "compression_ratio": 1.4747191011235956,
      "source": "whisper"
    },
    {
      "start": 474.765,
      "end": 477.525,
      "text": "最终生成的样子确实是一只狗",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.24606865776909723,
      "no_speech_prob": 4.710152406994794e-11,
      "compression_ratio": 1.4747191011235956,
      "source": "whisper"
    },
    {
      "start": 477.525,
      "end": 480.445,
      "text": "他没有说最后变成人活还是怎么样的",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.24606865776909723,
      "no_speech_prob": 4.710152406994794e-11,
      "compression_ratio": 1.4747191011235956,
      "source": "whisper"
    },
    {
      "start": 480.445,
      "end": 482.725,
      "text": "但是他这个样子就勉强动动嘴",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.24606865776909723,
      "no_speech_prob": 4.710152406994794e-11,
      "compression_ratio": 1.4747191011235956,
      "source": "whisper"
    },
    {
      "start": 482.725,
      "end": 483.925,
      "text": "如果用真人我之前试着",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.24606865776909723,
      "no_speech_prob": 4.710152406994794e-11,
      "compression_ratio": 1.4747191011235956,
      "source": "whisper"
    },
    {
      "start": 483.925,
      "end": 484.885,
      "text": "本次测试没有试过来",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.24606865776909723,
      "no_speech_prob": 4.710152406994794e-11,
      "compression_ratio": 1.4747191011235956,
      "source": "whisper"
    },
    {
      "start": 484.885,
      "end": 485.965,
      "text": "几秒钟之后",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.24606865776909723,
      "no_speech_prob": 4.710152406994794e-11,
      "compression_ratio": 1.4747191011235956,
      "source": "whisper"
    },
    {
      "start": 485.965,
      "end": 487.325,
      "text": "一两秒立马就变形了",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.24606865776909723,
      "no_speech_prob": 4.710152406994794e-11,
      "compression_ratio": 1.4747191011235956,
      "source": "whisper"
    },
    {
      "start": 487.325,
      "end": 489.325,
      "text": "具体生成的视频的样子",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.24606865776909723,
      "no_speech_prob": 4.710152406994794e-11,
      "compression_ratio": 1.4747191011235956,
      "source": "whisper"
    },
    {
      "start": 489.325,
      "end": 490.325,
      "text": "我们简单看一下",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.24606865776909723,
      "no_speech_prob": 4.710152406994794e-11,
      "compression_ratio": 1.4747191011235956,
      "source": "whisper"
    },
    {
      "start": 490.325,
      "end": 491.325,
      "text": "中流版的",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.24606865776909723,
      "no_speech_prob": 4.710152406994794e-11,
      "compression_ratio": 1.4747191011235956,
      "source": "whisper"
    },
    {
      "start": 491.325,
      "end": 492.005,
      "text": "这是美女",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.24606865776909723,
      "no_speech_prob": 4.710152406994794e-11,
      "compression_ratio": 1.4747191011235956,
      "source": "whisper"
    },
    {
      "start": 492.005,
      "end": 492.565,
      "text": "运动美女",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.24606865776909723,
      "no_speech_prob": 4.710152406994794e-11,
      "compression_ratio": 1.4747191011235956,
      "source": "whisper"
    },
    {
      "start": 492.565,
      "end": 493.285,
      "text": "还得车库的",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.24606865776909723,
      "no_speech_prob": 4.710152406994794e-11,
      "compression_ratio": 1.4747191011235956,
      "source": "whisper"
    },
    {
      "start": 493.285,
      "end": 494.245,
      "text": "是两阶段版本的两个",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.24606865776909723,
      "no_speech_prob": 4.710152406994794e-11,
      "compression_ratio": 1.4747191011235956,
      "source": "whisper"
    },
    {
      "start": 494.245,
      "end": 495.165,
      "text": "很明显动作",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.24606865776909723,
      "no_speech_prob": 4.710152406994794e-11,
      "compression_ratio": 1.4747191011235956,
      "source": "whisper"
    },
    {
      "start": 495.165,
      "end": 497.685,
      "text": "他的样子变得就更加活跃一点",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.24606865776909723,
      "no_speech_prob": 4.710152406994794e-11,
      "compression_ratio": 1.4747191011235956,
      "source": "whisper"
    },
    {
      "start": 497.685,
      "end": 498.325,
      "text": "动作也有了",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.24606865776909723,
      "no_speech_prob": 4.710152406994794e-11,
      "compression_ratio": 1.4747191011235956,
      "source": "whisper"
    },
    {
      "start": 498.325,
      "end": 498.925,
      "text": "是实验室的",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.24606865776909723,
      "no_speech_prob": 4.710152406994794e-11,
      "compression_ratio": 1.4747191011235956,
      "source": "whisper"
    },
    {
      "start": 498.925,
      "end": 500.005,
      "text": "一个阿三在这白话",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.24606865776909723,
      "no_speech_prob": 4.710152406994794e-11,
      "compression_ratio": 1.4747191011235956,
      "source": "whisper"
    },
    {
      "start": 500.005,
      "end": 501.325,
      "text": "就像是一个泰国人他在说话",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16360916864304315,
      "no_speech_prob": 4.517385424063214e-11,
      "compression_ratio": 1.464102564102564,
      "source": "whisper"
    },
    {
      "start": 501.325,
      "end": 503.685,
      "text": "你看他脸上这种雀斑这种表情",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16360916864304315,
      "no_speech_prob": 4.517385424063214e-11,
      "compression_ratio": 1.464102564102564,
      "source": "whisper"
    },
    {
      "start": 503.685,
      "end": 505.545,
      "text": "我们能看到他确实挺像真人的",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16360916864304315,
      "no_speech_prob": 4.517385424063214e-11,
      "compression_ratio": 1.464102564102564,
      "source": "whisper"
    },
    {
      "start": 505.545,
      "end": 506.345,
      "text": "这个效果是不错的",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16360916864304315,
      "no_speech_prob": 4.517385424063214e-11,
      "compression_ratio": 1.464102564102564,
      "source": "whisper"
    },
    {
      "start": 506.345,
      "end": 508.205,
      "text": "综合评价我觉得就差很多了",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16360916864304315,
      "no_speech_prob": 4.517385424063214e-11,
      "compression_ratio": 1.464102564102564,
      "source": "whisper"
    },
    {
      "start": 508.205,
      "end": 509.845,
      "text": "首先速度相对慢",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16360916864304315,
      "no_speech_prob": 4.517385424063214e-11,
      "compression_ratio": 1.464102564102564,
      "source": "whisper"
    },
    {
      "start": 509.845,
      "end": 512.225,
      "text": "然后这个T14理解稍显差",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16360916864304315,
      "no_speech_prob": 4.517385424063214e-11,
      "compression_ratio": 1.464102564102564,
      "source": "whisper"
    },
    {
      "start": 512.225,
      "end": 513.665,
      "text": "然后身份保持也不太行",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16360916864304315,
      "no_speech_prob": 4.517385424063214e-11,
      "compression_ratio": 1.464102564102564,
      "source": "whisper"
    },
    {
      "start": 513.665,
      "end": 514.625,
      "text": "所以我觉得是这样",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16360916864304315,
      "no_speech_prob": 4.517385424063214e-11,
      "compression_ratio": 1.464102564102564,
      "source": "whisper"
    },
    {
      "start": 514.625,
      "end": 515.765,
      "text": "这个视频生成能力",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16360916864304315,
      "no_speech_prob": 4.517385424063214e-11,
      "compression_ratio": 1.464102564102564,
      "source": "whisper"
    },
    {
      "start": 515.765,
      "end": 516.885,
      "text": "目前LTX模型",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16360916864304315,
      "no_speech_prob": 4.517385424063214e-11,
      "compression_ratio": 1.464102564102564,
      "source": "whisper"
    },
    {
      "start": 516.885,
      "end": 518.485,
      "text": "如果用一些视频里边的匕骤",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16360916864304315,
      "no_speech_prob": 4.517385424063214e-11,
      "compression_ratio": 1.464102564102564,
      "source": "whisper"
    },
    {
      "start": 518.485,
      "end": 519.965,
      "text": "或者进行一些原型的",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16360916864304315,
      "no_speech_prob": 4.517385424063214e-11,
      "compression_ratio": 1.464102564102564,
      "source": "whisper"
    },
    {
      "start": 519.965,
      "end": 521.605,
      "text": "比如展示一下你想要的东西",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16360916864304315,
      "no_speech_prob": 4.517385424063214e-11,
      "compression_ratio": 1.464102564102564,
      "source": "whisper"
    },
    {
      "start": 521.605,
      "end": 522.565,
      "text": "但是直接可能不行",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16360916864304315,
      "no_speech_prob": 4.517385424063214e-11,
      "compression_ratio": 1.464102564102564,
      "source": "whisper"
    },
    {
      "start": 522.565,
      "end": 524.025,
      "text": "我试了很多次全完蛋了",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16360916864304315,
      "no_speech_prob": 4.517385424063214e-11,
      "compression_ratio": 1.464102564102564,
      "source": "whisper"
    },
    {
      "start": 524.025,
      "end": 525.685,
      "text": "下一部分世界理解部分",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16360916864304315,
      "no_speech_prob": 4.517385424063214e-11,
      "compression_ratio": 1.464102564102564,
      "source": "whisper"
    },
    {
      "start": 525.685,
      "end": 526.925,
      "text": "这个世界理解部分呢",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16360916864304315,
      "no_speech_prob": 4.517385424063214e-11,
      "compression_ratio": 1.464102564102564,
      "source": "whisper"
    },
    {
      "start": 526.925,
      "end": 527.685,
      "text": "这个模型是这样的",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16360916864304315,
      "no_speech_prob": 4.517385424063214e-11,
      "compression_ratio": 1.464102564102564,
      "source": "whisper"
    },
    {
      "start": 527.685,
      "end": 528.705,
      "text": "这个模型带V的",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.16360916864304315,
      "no_speech_prob": 4.517385424063214e-11,
      "compression_ratio": 1.464102564102564,
      "source": "whisper"
    },
    {
      "start": 528.705,
      "end": 530.005,
      "text": "Vision 7B一般的模型",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14722086588541666,
      "no_speech_prob": 4.3448619702068925e-11,
      "compression_ratio": 1.4485714285714286,
      "source": "whisper"
    },
    {
      "start": 530.005,
      "end": 530.965,
      "text": "专门是理解图片",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14722086588541666,
      "no_speech_prob": 4.3448619702068925e-11,
      "compression_ratio": 1.4485714285714286,
      "source": "whisper"
    },
    {
      "start": 530.965,
      "end": 531.985,
      "text": "第一个是代码识别",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14722086588541666,
      "no_speech_prob": 4.3448619702068925e-11,
      "compression_ratio": 1.4485714285714286,
      "source": "whisper"
    },
    {
      "start": 531.985,
      "end": 533.385,
      "text": "第二个是一些图表识别",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14722086588541666,
      "no_speech_prob": 4.3448619702068925e-11,
      "compression_ratio": 1.4485714285714286,
      "source": "whisper"
    },
    {
      "start": 533.385,
      "end": 534.765,
      "text": "第三个是界面",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14722086588541666,
      "no_speech_prob": 4.3448619702068925e-11,
      "compression_ratio": 1.4485714285714286,
      "source": "whisper"
    },
    {
      "start": 534.765,
      "end": 536.505,
      "text": "这是一个模拟的计算机编程界面",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14722086588541666,
      "no_speech_prob": 4.3448619702068925e-11,
      "compression_ratio": 1.4485714285714286,
      "source": "whisper"
    },
    {
      "start": 536.505,
      "end": 539.345,
      "text": "然后还有一个识别图片",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14722086588541666,
      "no_speech_prob": 4.3448619702068925e-11,
      "compression_ratio": 1.4485714285714286,
      "source": "whisper"
    },
    {
      "start": 539.345,
      "end": 541.085,
      "text": "还是依旧是用这个狗为识别的",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14722086588541666,
      "no_speech_prob": 4.3448619702068925e-11,
      "compression_ratio": 1.4485714285714286,
      "source": "whisper"
    },
    {
      "start": 541.085,
      "end": 541.705,
      "text": "本来说是个人",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14722086588541666,
      "no_speech_prob": 4.3448619702068925e-11,
      "compression_ratio": 1.4485714285714286,
      "source": "whisper"
    },
    {
      "start": 541.705,
      "end": 542.825,
      "text": "然后让它识别",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14722086588541666,
      "no_speech_prob": 4.3448619702068925e-11,
      "compression_ratio": 1.4485714285714286,
      "source": "whisper"
    },
    {
      "start": 542.825,
      "end": 544.405,
      "text": "确实最后输出了",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14722086588541666,
      "no_speech_prob": 4.3448619702068925e-11,
      "compression_ratio": 1.4485714285714286,
      "source": "whisper"
    },
    {
      "start": 544.405,
      "end": 544.905,
      "text": "没有人物",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14722086588541666,
      "no_speech_prob": 4.3448619702068925e-11,
      "compression_ratio": 1.4485714285714286,
      "source": "whisper"
    },
    {
      "start": 544.905,
      "end": 545.665,
      "text": "只有一只哈布狗",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14722086588541666,
      "no_speech_prob": 4.3448619702068925e-11,
      "compression_ratio": 1.4485714285714286,
      "source": "whisper"
    },
    {
      "start": 545.665,
      "end": 546.845,
      "text": "这个相当厉害",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14722086588541666,
      "no_speech_prob": 4.3448619702068925e-11,
      "compression_ratio": 1.4485714285714286,
      "source": "whisper"
    },
    {
      "start": 546.845,
      "end": 549.225,
      "text": "不会遵从人的这种强制的复杂图片",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14722086588541666,
      "no_speech_prob": 4.3448619702068925e-11,
      "compression_ratio": 1.4485714285714286,
      "source": "whisper"
    },
    {
      "start": 549.225,
      "end": 550.285,
      "text": "需要识别的时候",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14722086588541666,
      "no_speech_prob": 4.3448619702068925e-11,
      "compression_ratio": 1.4485714285714286,
      "source": "whisper"
    },
    {
      "start": 550.285,
      "end": 552.405,
      "text": "它在统计数字的上面",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14722086588541666,
      "no_speech_prob": 4.3448619702068925e-11,
      "compression_ratio": 1.4485714285714286,
      "source": "whisper"
    },
    {
      "start": 552.405,
      "end": 553.945,
      "text": "确实不太主意",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14722086588541666,
      "no_speech_prob": 4.3448619702068925e-11,
      "compression_ratio": 1.4485714285714286,
      "source": "whisper"
    },
    {
      "start": 553.945,
      "end": 554.745,
      "text": "双图对比的时候",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14722086588541666,
      "no_speech_prob": 4.3448619702068925e-11,
      "compression_ratio": 1.4485714285714286,
      "source": "whisper"
    },
    {
      "start": 554.745,
      "end": 556.305,
      "text": "天鹅宝和在安风景",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.14722086588541666,
      "no_speech_prob": 4.3448619702068925e-11,
      "compression_ratio": 1.4485714285714286,
      "source": "whisper"
    },
    {
      "start": 556.305,
      "end": 558.165,
      "text": "他说是跟咖啡馆有关",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 558.165,
      "end": 558.865,
      "text": "有点偏差",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 558.865,
      "end": 560.025,
      "text": "是具体的一些结果",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 560.025,
      "end": 562.405,
      "text": "那么描述比如说场景",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 562.405,
      "end": 563.085,
      "text": "什么样子的",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 563.085,
      "end": 563.865,
      "text": "主体什么样的",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 563.865,
      "end": 564.265,
      "text": "大楼",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 564.265,
      "end": 565.445,
      "text": "包括毛树人物",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 565.445,
      "end": 566.985,
      "text": "这边他说了图片只有一只哈布狗",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 566.985,
      "end": 567.745,
      "text": "有一个分析图表",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 567.745,
      "end": 569.325,
      "text": "然后他给出了具体年月日",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 569.325,
      "end": 570.825,
      "text": "最高值平均值等等等等的",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 570.825,
      "end": 571.525,
      "text": "识别出来了",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 571.525,
      "end": 572.905,
      "text": "然后代码理解",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 572.905,
      "end": 574.825,
      "text": "他理解出来的相关的内容",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 574.825,
      "end": 575.405,
      "text": "贝波那奇数列",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 575.405,
      "end": 575.885,
      "text": "数东西",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 575.885,
      "end": 576.925,
      "text": "然后他数列大堆",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 576.925,
      "end": 577.725,
      "text": "这个数彩的花椒",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 577.725,
      "end": 578.205,
      "text": "分析图片",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 578.205,
      "end": 578.905,
      "text": "然后分析相同点",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 578.905,
      "end": 579.245,
      "text": "不同点",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 579.245,
      "end": 580.245,
      "text": "他说了包括哈维胡",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 580.245,
      "end": 580.925,
      "text": "当然刚才看了",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 580.925,
      "end": 581.265,
      "text": "这是风景",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15822289360894098,
      "no_speech_prob": 3.912739229838813e-11,
      "compression_ratio": 1.4869565217391305,
      "source": "whisper"
    },
    {
      "start": 581.265,
      "end": 581.865,
      "text": "最终结论吧",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15156929863823784,
      "no_speech_prob": 4.161698763383015e-11,
      "compression_ratio": 1.3144329896907216,
      "source": "whisper"
    },
    {
      "start": 581.865,
      "end": 584.205,
      "text": "首先实时对话这个是非常可用的",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15156929863823784,
      "no_speech_prob": 4.161698763383015e-11,
      "compression_ratio": 1.3144329896907216,
      "source": "whisper"
    },
    {
      "start": 584.205,
      "end": 585.125,
      "text": "流畅然后准确",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15156929863823784,
      "no_speech_prob": 4.161698763383015e-11,
      "compression_ratio": 1.3144329896907216,
      "source": "whisper"
    },
    {
      "start": 585.125,
      "end": 586.865,
      "text": "视觉理解是非常棒的",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15156929863823784,
      "no_speech_prob": 4.161698763383015e-11,
      "compression_ratio": 1.3144329896907216,
      "source": "whisper"
    },
    {
      "start": 586.865,
      "end": 587.325,
      "text": "大家看到",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15156929863823784,
      "no_speech_prob": 4.161698763383015e-11,
      "compression_ratio": 1.3144329896907216,
      "source": "whisper"
    },
    {
      "start": 587.325,
      "end": 588.985,
      "text": "它能准确识别数字啊",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15156929863823784,
      "no_speech_prob": 4.161698763383015e-11,
      "compression_ratio": 1.3144329896907216,
      "source": "whisper"
    },
    {
      "start": 588.985,
      "end": 589.865,
      "text": "C些图柄什么的",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15156929863823784,
      "no_speech_prob": 4.161698763383015e-11,
      "compression_ratio": 1.3144329896907216,
      "source": "whisper"
    },
    {
      "start": 589.865,
      "end": 590.665,
      "text": "外界的问题不大",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15156929863823784,
      "no_speech_prob": 4.161698763383015e-11,
      "compression_ratio": 1.3144329896907216,
      "source": "whisper"
    },
    {
      "start": 590.665,
      "end": 591.345,
      "text": "然后声音克隆",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15156929863823784,
      "no_speech_prob": 4.161698763383015e-11,
      "compression_ratio": 1.3144329896907216,
      "source": "whisper"
    },
    {
      "start": 591.345,
      "end": 592.125,
      "text": "刚才大家听到",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15156929863823784,
      "no_speech_prob": 4.161698763383015e-11,
      "compression_ratio": 1.3144329896907216,
      "source": "whisper"
    },
    {
      "start": 592.125,
      "end": 592.905,
      "text": "简直是完美",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15156929863823784,
      "no_speech_prob": 4.161698763383015e-11,
      "compression_ratio": 1.3144329896907216,
      "source": "whisper"
    },
    {
      "start": 592.905,
      "end": 593.885,
      "text": "这个图像生成呢",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15156929863823784,
      "no_speech_prob": 4.161698763383015e-11,
      "compression_ratio": 1.3144329896907216,
      "source": "whisper"
    },
    {
      "start": 593.885,
      "end": 594.965,
      "text": "等于是速度稍微慢",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15156929863823784,
      "no_speech_prob": 4.161698763383015e-11,
      "compression_ratio": 1.3144329896907216,
      "source": "whisper"
    },
    {
      "start": 594.965,
      "end": 595.785,
      "text": "克隆震灵来提升",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15156929863823784,
      "no_speech_prob": 4.161698763383015e-11,
      "compression_ratio": 1.3144329896907216,
      "source": "whisper"
    },
    {
      "start": 595.785,
      "end": 597.545,
      "text": "视觉模型的差距非常大",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15156929863823784,
      "no_speech_prob": 4.161698763383015e-11,
      "compression_ratio": 1.3144329896907216,
      "source": "whisper"
    },
    {
      "start": 597.545,
      "end": 599.565,
      "text": "只能生成一些基本的内容",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15156929863823784,
      "no_speech_prob": 4.161698763383015e-11,
      "compression_ratio": 1.3144329896907216,
      "source": "whisper"
    },
    {
      "start": 599.565,
      "end": 600.565,
      "text": "给人用氛围感",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15156929863823784,
      "no_speech_prob": 4.161698763383015e-11,
      "compression_ratio": 1.3144329896907216,
      "source": "whisper"
    },
    {
      "start": 600.565,
      "end": 601.285,
      "text": "人脸一致性",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15156929863823784,
      "no_speech_prob": 4.161698763383015e-11,
      "compression_ratio": 1.3144329896907216,
      "source": "whisper"
    },
    {
      "start": 601.285,
      "end": 602.105,
      "text": "基本上不可用",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15156929863823784,
      "no_speech_prob": 4.161698763383015e-11,
      "compression_ratio": 1.3144329896907216,
      "source": "whisper"
    },
    {
      "start": 602.105,
      "end": 605.025,
      "text": "这个MacBook Pro M564G配置上",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15156929863823784,
      "no_speech_prob": 4.161698763383015e-11,
      "compression_ratio": 1.3144329896907216,
      "source": "whisper"
    },
    {
      "start": 605.025,
      "end": 606.625,
      "text": "然后小而快的模型",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15156929863823784,
      "no_speech_prob": 4.161698763383015e-11,
      "compression_ratio": 1.3144329896907216,
      "source": "whisper"
    },
    {
      "start": 606.625,
      "end": 607.585,
      "text": "还是比较有价值",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.15156929863823784,
      "no_speech_prob": 4.161698763383015e-11,
      "compression_ratio": 1.3144329896907216,
      "source": "whisper"
    },
    {
      "start": 607.585,
      "end": 609.565,
      "text": "相信随着苹果架构本身的它的完善",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.08322069596271126,
      "no_speech_prob": 1.116712538412612e-11,
      "compression_ratio": 1.0344827586206897,
      "source": "whisper"
    },
    {
      "start": 609.565,
      "end": 611.445,
      "text": "相关配置环境的完善",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.08322069596271126,
      "no_speech_prob": 1.116712538412612e-11,
      "compression_ratio": 1.0344827586206897,
      "source": "whisper"
    },
    {
      "start": 611.445,
      "end": 614.805,
      "text": "我觉得它的提升应该还是会很迅速的",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.08322069596271126,
      "no_speech_prob": 1.116712538412612e-11,
      "compression_ratio": 1.0344827586206897,
      "source": "whisper"
    }
  ],
  "transcript_selection": {
    "schema": "videonote.transcript_selection.v1",
    "status": "SELECTED",
    "requested_language": "zh",
    "duration": 614.805313,
    "reference_path": null,
    "reference_kind": null,
    "reference_language": null,
    "selected_source": "whisper_segmented",
    "decision_reason": "source_unavailable_or_untrusted_whisper_passed",
    "whisper_rejected": false,
    "source_profile": {
      "segment_count": 0,
      "duration_coverage": 0.0,
      "usable_full": false,
      "usable_partial": false,
      "hallucination_region_count": 0,
      "hallucination_regions": [],
      "language": {
        "requested": "zh",
        "matched": false
      }
    },
    "whisper_profile": {
      "segment_count": 374,
      "multilingual": false,
      "first_start": 0.0,
      "final_end": 614.805,
      "duration_coverage": 0.999999,
      "maximum_gap_seconds": 2.22,
      "inverted_count": 0,
      "language": {
        "requested": "zh",
        "matched": true,
        "latin": 190,
        "han": 3493,
        "kana": 0,
        "hangul": 0
      },
      "hallucination_region_count": 0,
      "hallucination_regions": [],
      "usable_full": true,
      "usable_partial": true
    },
    "comparison": {
      "window_seconds": 30.0,
      "compared_window_count": 0,
      "median_score": 0.0,
      "divergent_window_count": 0,
      "divergent_ratio": 0.0,
      "divergent_windows": []
    },
    "timestamp_alignment": {
      "applied": false,
      "estimated_offset_seconds": 0.0,
      "matched_segment_count": 0,
      "match_scores": []
    },
    "caption_timing_repair": null,
    "filled_source_profile": null,
    "whisper_duration_guard": {
      "media_duration": 614.805,
      "rejected_segment_count": 0,
      "rejected_segments": []
    }
  }
}