{"text": "Hello 各位开发者朋友欢迎收听GitCoverty今天我们要带你快速掌握GitHub Trending上最火红的开源专案让你不错过任何值得关注的技术动态今天我们要深度解析GitHub上的一个热门专案PDF Inspector一个用RUST打造的高效能PDF处理引擎它能智慧地区分扫描档与文字档让你避免对超过50%的文件进行昂贵又耗时的OCR处理先来问一个问题当你的系统需要处理海量的PDF文件到底有多少成本和时间是浪费在根本不必要的处理上答案可能会让你吓一跳现在有一个叫做PDF Inspector的开源专案它就想透过智慧分类来帮你找到一个超有效率的解答根据这个专案的分析竟然有高达54%的PDF根本不需要用到昂贵的光学资源辨识这代表什么代表你一半以上的处理成本都有机会省下来这个PDF Inspector专案是由FireCrow团队开发而且是用Rust语言写成的高效能函试库它的定位很清楚就是要成为PDF处理流程里那个最聪明的守门员它能做什么呢它可以快速检测分类PDF还能从里面抓出结构化的文字这个专案在开源社群爆红的速度非常快目前已经累积超过8000颗Github星星光是今天一天就增加了快要1700颗这就充分说明了开发者社群对它的解决方案有多么买单它的主要目标客群就是那些需要大规模高效能处理文件的后端开发者数据工程师还有AI领域的专家们一直以来开发者在处理PDF的时候都卡在一个两栏的困境里简单来说PDF分成两大类第一种是原生文字型它的文字资讯本来就欠在档案里面另一种是扫描型档案本质上就是一张大图片你必须透过又花时间又花钱的光学资源辨识也就是OCR才有办法把文字读出来那传统的做法是怎么样呢为了保险起见大家常常不管三七二十一把所有PDF全部丢给OCR服务处理这种一刀切的做法不只让云端服务的费用暴增也造成了严重的处理延迟而PDF Inspector这个专案切入的正是这个痛点它不做那种什么都包的大杂烩功能而是专心做好第一步用最快的速度判断出这份PDF到底需不需要跑OCRPDF Inspector之所以这么有效率关键就在它精巧的架构设计你可以把它想象成医院的减伤分类站当一份PDF文件送进来它不会马上做全身健康检查相反地会先交给一个速度超快的侦测器模组这个侦测器很聪明它不读完整个档案只抽样检查PDF里面那些代表文字或图像的操作指令光是靠着分析这些指令它就能在短短10到50毫秒内判断出这份文件是文字型扫描型还是混合型这个速度就是它甩开其他工具的关键如果判断是文字型档案才会被交给下一步的提取器模组来做深度的处理整个流程最精彩的设计就是PDF档案从头到尾只会被读取一次它的分析结果会由侦测和提取这两个模组共享这样就彻底避免了重复读取和运算的浪费这种单次载入多方共享的架构正是它能够又快又准的核心秘密那在实际应用上PDF Inspector的场景就非常清楚了首先想象一下那些需要处理大量报告发票或法律文件的后端系统开发者可以把PDF Inspector当作整个流程的第一关它就像一个智慧路由器可以快速的把文字型PDF指向本地的服务来快速提取只有那些真正的扫描件才需要送去昂贵的云端OCR这样一来成本和延迟都能大幅降低再来在AI HAN REC也就是检索增强生成的应用里数据科学家可以用它快速把研究论文这类的PDF转成结构清楚的Markdown格式直接拿来当作大型语言模型的训练资料更酷的是它还提供了Web Assembly版本这代表什么这代表前端开发者可以直接在浏览器里就分析使用者上传的PDF完全不需要跟伺服器来回沟通既能提升效能又能保护使用者隐私而且因为专案提供了Python Node.js这些主流语言的绑定上手的门槛也相当低社群对PDF Inspector的反应普遍都非常好大家称赞的点主要都集中在它那快到下人的处理速度还有完全可以在本地端执行的能力很多开发者都说光是它提供的需要OCR的页面列表这个功能就帮他们省下了大笔的云端服务账单社群也特别提到它输出的Markdown品质很好非常适合直接喂给大型语言模型目前还没有看到什么明显的负评不过倒是有一些很务实的提醒比方说这个专案本身不包含OCR功能所以如果遇到扫描文件还是得搭配其他的引擎才能处理如果我们把眼光放远一点来看PDF Inspector的出现其实正在挑战整个业界处理文件的预设思维它所提倡的分类优先模式会促使开发者从过去那种暴力破解式的OCR处理转向更精细也更有成本效益的智慧化流程不过它的潜在风险也不能忽视首先这个专案高度依赖底层的一个叫做LobDF的函事库所以只要LobDF有任何限制都会直接影响到它其次它的分类和版面分析大量采用了启发式演算法这代表说当它遇到那种特别复杂或是不照规矩来的版面时还是有可能会判断错误最后它清楚的功能边界也代表使用者必须自己去整合第三方的OCR服务才能打造出一个完整的解决方案总结来说PDF Inspector并不是又一个单纯的PDF转档工具它更像是一个为了现代化文件处理流程所设计的智慧调度核心它的价值并不是要取代OCR而是要让OCR的使用变得更聪明更有效率对于任何需要大规模处理PDF的开发者或组织来说这个专案最值得关注的地方就是它提供了一个绝佳的机会让你用极低的整合成本去重新检视并且优化你整个系统里那个最花钱的环节以上就是今天的GitCoverty希望这些GitHub Trending专案能为你的开发工作带来灵感以上这些GitHub Report请见下方资讯栏如果你喜欢这个节目别忘了订阅我们的频道我们明天同一时间再见", "segments": [{"id": 0, "seek": 0, "start": 0.0, "end": 1.36, "text": "Hello 各位开发者朋友", "tokens": [50365, 15947, 220, 22073, 18937, 28926, 12444, 19828, 50433], "temperature": 0, "avg_logprob": -0.14136800155263815, "compression_ratio": 1.2189973614775726, "no_speech_prob": 1.8680930066738988e-11}, {"id": 1, "seek": 0, "start": 1.36, "end": 2.64, "text": "欢迎收听GitCoverty", "tokens": [50433, 28566, 17699, 18681, 31022, 38, 270, 34, 3570, 874, 50497], "temperature": 0, "avg_logprob": -0.14136800155263815, "compression_ratio": 1.2189973614775726, "no_speech_prob": 1.8680930066738988e-11}, {"id": 2, "seek": 0, "start": 2.64, "end": 5.36, "text": "今天我们要带你快速掌握GitHub Trending上", "tokens": [50497, 12074, 15003, 4275, 4845, 99, 2166, 10251, 31217, 6900, 234, 11673, 94, 38, 270, 21150, 37417, 278, 5708, 50633], "temperature": 0, "avg_logprob": -0.14136800155263815, "compression_ratio": 1.2189973614775726, "no_speech_prob": 1.8680930066738988e-11}, {"id": 3, "seek": 0, "start": 5.36, "end": 7.04, "text": "最火红的开源专案", "tokens": [50633, 8661, 26171, 16853, 95, 1546, 18937, 47402, 940, 241, 28899, 50717], "temperature": 0, "avg_logprob": -0.14136800155263815, "compression_ratio": 1.2189973614775726, "no_speech_prob": 1.8680930066738988e-11}, {"id": 4, "seek": 0, "start": 7.04, "end": 9.96, "text": "让你不错过任何值得关注的技术动态", "tokens": [50717, 33650, 2166, 1960, 29900, 16866, 26443, 11906, 40242, 5916, 28053, 26432, 1546, 32502, 1474, 107, 34961, 3757, 223, 50863], "temperature": 0, "avg_logprob": -0.14136800155263815, "compression_ratio": 1.2189973614775726, "no_speech_prob": 1.8680930066738988e-11}, {"id": 5, "seek": 0, "start": 9.96, "end": 13.76, "text": "今天我们要深度解析GitHub上的一个热门专案", "tokens": [50863, 12074, 15003, 4275, 24043, 13127, 17278, 7360, 238, 38, 270, 21150, 5708, 1546, 20182, 23661, 255, 8259, 101, 940, 241, 28899, 51053], "temperature": 0, "avg_logprob": -0.14136800155263815, "compression_ratio": 1.2189973614775726, "no_speech_prob": 1.8680930066738988e-11}, {"id": 6, "seek": 0, "start": 13.76, "end": 15.200000000000001, "text": "PDF Inspector", "tokens": [51053, 17349, 37, 33402, 51125], "temperature": 0, "avg_logprob": -0.14136800155263815, "compression_ratio": 1.2189973614775726, "no_speech_prob": 1.8680930066738988e-11}, {"id": 7, "seek": 0, "start": 15.200000000000001, "end": 18.6, "text": "一个用RUST打造的高效能PDF处理引擎", "tokens": [51125, 20182, 9254, 49, 21536, 12467, 37583, 1546, 12979, 43076, 8225, 17349, 37, 1787, 226, 13876, 41301, 13167, 236, 51295], "temperature": 0, "avg_logprob": -0.14136800155263815, "compression_ratio": 1.2189973614775726, "no_speech_prob": 1.8680930066738988e-11}, {"id": 8, "seek": 0, "start": 18.6, "end": 21.240000000000002, "text": "它能智慧地区分扫描档与文字档", "tokens": [51295, 11284, 8225, 5094, 118, 12358, 100, 10928, 9937, 118, 6627, 3416, 104, 11673, 237, 21416, 96, 940, 236, 17174, 22381, 21416, 96, 51427], "temperature": 0, "avg_logprob": -0.14136800155263815, "compression_ratio": 1.2189973614775726, "no_speech_prob": 1.8680930066738988e-11}, {"id": 9, "seek": 0, "start": 21.240000000000002, "end": 25.64, "text": "让你避免对超过50%的文件进行昂贵又耗时的OCR处理", "tokens": [51427, 33650, 2166, 3330, 123, 2347, 235, 8713, 19869, 16866, 2803, 4, 1546, 17174, 20485, 36700, 8082, 1431, 224, 18464, 113, 17047, 4450, 245, 15729, 1546, 46, 18547, 1787, 226, 13876, 51647], "temperature": 0, "avg_logprob": -0.14136800155263815, "compression_ratio": 1.2189973614775726, "no_speech_prob": 1.8680930066738988e-11}, {"id": 10, "seek": 0, "start": 25.64, "end": 26.76, "text": "先来问一个问题", "tokens": [51647, 10108, 6912, 22064, 20182, 34069, 51703], "temperature": 0, "avg_logprob": -0.14136800155263815, "compression_ratio": 1.2189973614775726, "no_speech_prob": 1.8680930066738988e-11}, {"id": 11, "seek": 0, "start": 26.76, "end": 29.64, "text": "当你的系统需要处理海量的PDF文件", "tokens": [51703, 16233, 18961, 25368, 10115, 253, 35748, 1787, 226, 13876, 20078, 26748, 1546, 17349, 37, 17174, 20485, 51847], "temperature": 0, "avg_logprob": -0.14136800155263815, "compression_ratio": 1.2189973614775726, "no_speech_prob": 1.8680930066738988e-11}, {"id": 12, "seek": 2964, "start": 29.64, "end": 31.44, "text": "到底有多少成本和时间", "tokens": [50365, 33883, 2412, 41492, 11336, 8802, 12565, 44848, 50455], "temperature": 0, "avg_logprob": -0.08057301567200033, "compression_ratio": 1.272463768115942, "no_speech_prob": 1.1217572010169263e-11}, {"id": 13, "seek": 2964, "start": 31.44, "end": 33.58, "text": "是浪费在根本不必要的处理上", "tokens": [50455, 1541, 11038, 103, 18464, 117, 3581, 31337, 8802, 1960, 28531, 4275, 1546, 1787, 226, 13876, 5708, 50562], "temperature": 0, "avg_logprob": -0.08057301567200033, "compression_ratio": 1.272463768115942, "no_speech_prob": 1.1217572010169263e-11}, {"id": 14, "seek": 2964, "start": 33.58, "end": 35.160000000000004, "text": "答案可能会让你吓一跳", "tokens": [50562, 28223, 28899, 16657, 12949, 33650, 2166, 2392, 241, 2257, 29366, 50641], "temperature": 0, "avg_logprob": -0.08057301567200033, "compression_ratio": 1.272463768115942, "no_speech_prob": 1.1217572010169263e-11}, {"id": 15, "seek": 2964, "start": 35.160000000000004, "end": 38.4, "text": "现在有一个叫做PDF Inspector的开源专案", "tokens": [50641, 25040, 2412, 20182, 19855, 10907, 17349, 37, 33402, 1546, 18937, 47402, 940, 241, 28899, 50803], "temperature": 0, "avg_logprob": -0.08057301567200033, "compression_ratio": 1.272463768115942, "no_speech_prob": 1.1217572010169263e-11}, {"id": 16, "seek": 2964, "start": 38.4, "end": 40.02, "text": "它就想透过智慧分类", "tokens": [50803, 11284, 3111, 7093, 45808, 16866, 5094, 118, 12358, 100, 6627, 22113, 119, 50884], "temperature": 0, "avg_logprob": -0.08057301567200033, "compression_ratio": 1.272463768115942, "no_speech_prob": 1.1217572010169263e-11}, {"id": 17, "seek": 2964, "start": 40.02, "end": 42.72, "text": "来帮你找到一个超有效率的解答", "tokens": [50884, 6912, 4845, 106, 2166, 25085, 4511, 20182, 19869, 2412, 43076, 44866, 1546, 17278, 28223, 51019], "temperature": 0, "avg_logprob": -0.08057301567200033, "compression_ratio": 1.272463768115942, "no_speech_prob": 1.1217572010169263e-11}, {"id": 18, "seek": 2964, "start": 42.72, "end": 44.32, "text": "根据这个专案的分析", "tokens": [51019, 31337, 26075, 106, 15368, 940, 241, 28899, 1546, 6627, 7360, 238, 51099], "temperature": 0, "avg_logprob": -0.08057301567200033, "compression_ratio": 1.272463768115942, "no_speech_prob": 1.1217572010169263e-11}, {"id": 19, "seek": 2964, "start": 44.32, "end": 47.120000000000005, "text": "竟然有高达54%的PDF", "tokens": [51099, 11957, 253, 5823, 2412, 12979, 9830, 122, 19563, 4, 1546, 17349, 37, 51239], "temperature": 0, "avg_logprob": -0.08057301567200033, "compression_ratio": 1.272463768115942, "no_speech_prob": 1.1217572010169263e-11}, {"id": 20, "seek": 2964, "start": 47.120000000000005, "end": 50.040000000000006, "text": "根本不需要用到昂贵的光学资源辨识", "tokens": [51239, 31337, 8802, 1960, 35748, 9254, 4511, 1431, 224, 18464, 113, 1546, 20690, 29618, 5266, 226, 47402, 9830, 101, 5233, 228, 51385], "temperature": 0, "avg_logprob": -0.08057301567200033, "compression_ratio": 1.272463768115942, "no_speech_prob": 1.1217572010169263e-11}, {"id": 21, "seek": 2964, "start": 50.040000000000006, "end": 50.8, "text": "这代表什么", "tokens": [51385, 5562, 19105, 17571, 10440, 51423], "temperature": 0, "avg_logprob": -0.08057301567200033, "compression_ratio": 1.272463768115942, "no_speech_prob": 1.1217572010169263e-11}, {"id": 22, "seek": 2964, "start": 50.8, "end": 52.540000000000006, "text": "代表你一半以上的处理成本", "tokens": [51423, 19105, 17571, 2166, 2257, 30018, 29497, 1546, 1787, 226, 13876, 11336, 8802, 51510], "temperature": 0, "avg_logprob": -0.08057301567200033, "compression_ratio": 1.272463768115942, "no_speech_prob": 1.1217572010169263e-11}, {"id": 23, "seek": 2964, "start": 52.540000000000006, "end": 53.84, "text": "都有机会省下来", "tokens": [51510, 48121, 37960, 12949, 2862, 223, 4438, 6912, 51575], "temperature": 0, "avg_logprob": -0.08057301567200033, "compression_ratio": 1.272463768115942, "no_speech_prob": 1.1217572010169263e-11}, {"id": 24, "seek": 2964, "start": 53.84, "end": 55.78, "text": "这个PDF Inspector专案", "tokens": [51575, 15368, 17349, 37, 33402, 940, 241, 28899, 51672], "temperature": 0, "avg_logprob": -0.08057301567200033, "compression_ratio": 1.272463768115942, "no_speech_prob": 1.1217572010169263e-11}, {"id": 25, "seek": 2964, "start": 55.78, "end": 57.6, "text": "是由FireCrow团队开发", "tokens": [51672, 1541, 23786, 45651, 34, 1892, 3919, 95, 10034, 253, 18937, 28926, 51763], "temperature": 0, "avg_logprob": -0.08057301567200033, "compression_ratio": 1.272463768115942, "no_speech_prob": 1.1217572010169263e-11}, {"id": 26, "seek": 5760, "start": 57.6, "end": 60.52, "text": "而且是用Rust语言写成的高效能函试库", "tokens": [50365, 22942, 1541, 9254, 49, 381, 5233, 255, 12009, 5676, 247, 11336, 1546, 12979, 43076, 8225, 6336, 121, 5233, 243, 6346, 241, 50511], "temperature": 0, "avg_logprob": -0.10506588347414707, "compression_ratio": 1.2206303724928367, "no_speech_prob": 1.1706577547621055e-11}, {"id": 27, "seek": 5760, "start": 60.52, "end": 61.88, "text": "它的定位很清楚", "tokens": [50511, 45224, 12088, 11160, 4563, 38433, 50579], "temperature": 0, "avg_logprob": -0.10506588347414707, "compression_ratio": 1.2206303724928367, "no_speech_prob": 1.1706577547621055e-11}, {"id": 28, "seek": 5760, "start": 61.88, "end": 64.28, "text": "就是要成为PDF处理流程里", "tokens": [50579, 5620, 4275, 11336, 13992, 17349, 37, 1787, 226, 13876, 27854, 29649, 15759, 50699], "temperature": 0, "avg_logprob": -0.10506588347414707, "compression_ratio": 1.2206303724928367, "no_speech_prob": 1.1706577547621055e-11}, {"id": 29, "seek": 5760, "start": 64.28, "end": 65.84, "text": "那个最聪明的守门员", "tokens": [50699, 43083, 8661, 8171, 103, 11100, 1546, 48480, 8259, 101, 3606, 246, 50777], "temperature": 0, "avg_logprob": -0.10506588347414707, "compression_ratio": 1.2206303724928367, "no_speech_prob": 1.1706577547621055e-11}, {"id": 30, "seek": 5760, "start": 65.84, "end": 66.96000000000001, "text": "它能做什么呢", "tokens": [50777, 11284, 8225, 10907, 10440, 6240, 50833], "temperature": 0, "avg_logprob": -0.10506588347414707, "compression_ratio": 1.2206303724928367, "no_speech_prob": 1.1706577547621055e-11}, {"id": 31, "seek": 5760, "start": 66.96000000000001, "end": 68.26, "text": "它可以快速检测", "tokens": [50833, 11284, 6723, 10251, 31217, 16407, 222, 11038, 233, 50898], "temperature": 0, "avg_logprob": -0.10506588347414707, "compression_ratio": 1.2206303724928367, "no_speech_prob": 1.1706577547621055e-11}, {"id": 32, "seek": 5760, "start": 68.26, "end": 69.52, "text": "分类PDF", "tokens": [50898, 6627, 22113, 119, 17349, 37, 50961], "temperature": 0, "avg_logprob": -0.10506588347414707, "compression_ratio": 1.2206303724928367, "no_speech_prob": 1.1706577547621055e-11}, {"id": 33, "seek": 5760, "start": 69.52, "end": 71.8, "text": "还能从里面抓出结构化的文字", "tokens": [50961, 14852, 8225, 35630, 15759, 8833, 43069, 7781, 45641, 7360, 226, 23756, 1546, 17174, 22381, 51075], "temperature": 0, "avg_logprob": -0.10506588347414707, "compression_ratio": 1.2206303724928367, "no_speech_prob": 1.1706577547621055e-11}, {"id": 34, "seek": 5760, "start": 71.8, "end": 74.7, "text": "这个专案在开源社群爆红的速度非常快", "tokens": [51075, 15368, 940, 241, 28899, 3581, 18937, 47402, 27658, 7248, 97, 45827, 16853, 95, 1546, 31217, 13127, 14392, 10251, 51220], "temperature": 0, "avg_logprob": -0.10506588347414707, "compression_ratio": 1.2206303724928367, "no_speech_prob": 1.1706577547621055e-11}, {"id": 35, "seek": 5760, "start": 74.7, "end": 77.66, "text": "目前已经累积超过8000颗Github星星", "tokens": [51220, 39004, 49161, 43656, 8204, 107, 19869, 16866, 23, 1360, 12501, 245, 38, 355, 836, 20682, 20682, 51368], "temperature": 0, "avg_logprob": -0.10506588347414707, "compression_ratio": 1.2206303724928367, "no_speech_prob": 1.1706577547621055e-11}, {"id": 36, "seek": 5760, "start": 77.66, "end": 78.88, "text": "光是今天一天", "tokens": [51368, 20690, 1541, 12074, 2257, 6135, 51429], "temperature": 0, "avg_logprob": -0.10506588347414707, "compression_ratio": 1.2206303724928367, "no_speech_prob": 1.1706577547621055e-11}, {"id": 37, "seek": 5760, "start": 78.88, "end": 81.0, "text": "就增加了快要1700颗", "tokens": [51429, 3111, 24228, 252, 9990, 2289, 10251, 4275, 7773, 628, 12501, 245, 51535], "temperature": 0, "avg_logprob": -0.10506588347414707, "compression_ratio": 1.2206303724928367, "no_speech_prob": 1.1706577547621055e-11}, {"id": 38, "seek": 5760, "start": 81.0, "end": 82.32, "text": "这就充分说明了", "tokens": [51535, 5562, 3111, 2347, 227, 6627, 8090, 11100, 2289, 51601], "temperature": 0, "avg_logprob": -0.10506588347414707, "compression_ratio": 1.2206303724928367, "no_speech_prob": 1.1706577547621055e-11}, {"id": 39, "seek": 5760, "start": 82.32, "end": 85.2, "text": "开发者社群对它的解决方案有多么买单", "tokens": [51601, 18937, 28926, 12444, 27658, 7248, 97, 8713, 45224, 17278, 5676, 111, 9249, 28899, 2412, 6392, 6656, 2930, 108, 47446, 51745], "temperature": 0, "avg_logprob": -0.10506588347414707, "compression_ratio": 1.2206303724928367, "no_speech_prob": 1.1706577547621055e-11}, {"id": 40, "seek": 8520, "start": 85.2, "end": 88.08, "text": "它的主要目标客群就是那些需要大规模", "tokens": [50365, 45224, 13557, 4275, 11386, 162, 3921, 32316, 7248, 97, 5620, 4184, 13824, 35748, 3582, 6758, 226, 41908, 50509], "temperature": 0, "avg_logprob": -0.09122826128589864, "compression_ratio": 1.3231707317073171, "no_speech_prob": 1.4075023264947806e-11}, {"id": 41, "seek": 8520, "start": 88.08, "end": 90.36, "text": "高效能处理文件的后端开发者", "tokens": [50509, 12979, 43076, 8225, 1787, 226, 13876, 17174, 20485, 1546, 13547, 11957, 107, 18937, 28926, 12444, 50623], "temperature": 0, "avg_logprob": -0.09122826128589864, "compression_ratio": 1.3231707317073171, "no_speech_prob": 1.4075023264947806e-11}, {"id": 42, "seek": 8520, "start": 90.36, "end": 91.56, "text": "数据工程师", "tokens": [50623, 33188, 26075, 106, 23323, 29649, 29186, 50683], "temperature": 0, "avg_logprob": -0.09122826128589864, "compression_ratio": 1.3231707317073171, "no_speech_prob": 1.4075023264947806e-11}, {"id": 43, "seek": 8520, "start": 91.56, "end": 93.02000000000001, "text": "还有AI领域的专家们", "tokens": [50683, 35091, 48698, 12501, 228, 16262, 253, 1546, 940, 241, 5155, 9497, 50756], "temperature": 0, "avg_logprob": -0.09122826128589864, "compression_ratio": 1.3231707317073171, "no_speech_prob": 1.4075023264947806e-11}, {"id": 44, "seek": 8520, "start": 93.02000000000001, "end": 93.78, "text": "一直以来", "tokens": [50756, 34448, 3588, 6912, 50794], "temperature": 0, "avg_logprob": -0.09122826128589864, "compression_ratio": 1.3231707317073171, "no_speech_prob": 1.4075023264947806e-11}, {"id": 45, "seek": 8520, "start": 93.78, "end": 95.58, "text": "开发者在处理PDF的时候", "tokens": [50794, 18937, 28926, 12444, 3581, 1787, 226, 13876, 17349, 37, 49873, 50884], "temperature": 0, "avg_logprob": -0.09122826128589864, "compression_ratio": 1.3231707317073171, "no_speech_prob": 1.4075023264947806e-11}, {"id": 46, "seek": 8520, "start": 95.58, "end": 97.54, "text": "都卡在一个两栏的困境里", "tokens": [50884, 7182, 32681, 3581, 20182, 36257, 8987, 237, 1546, 3919, 108, 47885, 15759, 50982], "temperature": 0, "avg_logprob": -0.09122826128589864, "compression_ratio": 1.3231707317073171, "no_speech_prob": 1.4075023264947806e-11}, {"id": 47, "seek": 8520, "start": 97.54, "end": 98.52000000000001, "text": "简单来说", "tokens": [50982, 11249, 222, 47446, 6912, 8090, 51031], "temperature": 0, "avg_logprob": -0.09122826128589864, "compression_ratio": 1.3231707317073171, "no_speech_prob": 1.4075023264947806e-11}, {"id": 48, "seek": 8520, "start": 98.52000000000001, "end": 99.94, "text": "PDF分成两大类", "tokens": [51031, 17349, 37, 6627, 11336, 36257, 3582, 22113, 119, 51102], "temperature": 0, "avg_logprob": -0.09122826128589864, "compression_ratio": 1.3231707317073171, "no_speech_prob": 1.4075023264947806e-11}, {"id": 49, "seek": 8520, "start": 99.94, "end": 101.9, "text": "第一种是原生文字型", "tokens": [51102, 18049, 39810, 1541, 19683, 8244, 17174, 22381, 39823, 51200], "temperature": 0, "avg_logprob": -0.09122826128589864, "compression_ratio": 1.3231707317073171, "no_speech_prob": 1.4075023264947806e-11}, {"id": 50, "seek": 8520, "start": 101.9, "end": 104.76, "text": "它的文字资讯本来就欠在档案里面", "tokens": [51200, 45224, 17174, 22381, 5266, 226, 7422, 107, 8802, 6912, 3111, 5988, 254, 3581, 21416, 96, 28899, 15759, 8833, 51343], "temperature": 0, "avg_logprob": -0.09122826128589864, "compression_ratio": 1.3231707317073171, "no_speech_prob": 1.4075023264947806e-11}, {"id": 51, "seek": 8520, "start": 104.76, "end": 105.88, "text": "另一种是扫描型", "tokens": [51343, 22762, 2257, 39810, 1541, 3416, 104, 11673, 237, 39823, 51399], "temperature": 0, "avg_logprob": -0.09122826128589864, "compression_ratio": 1.3231707317073171, "no_speech_prob": 1.4075023264947806e-11}, {"id": 52, "seek": 8520, "start": 105.88, "end": 107.88, "text": "档案本质上就是一张大图片", "tokens": [51399, 21416, 96, 28899, 8802, 18464, 101, 5708, 5620, 2257, 44059, 3582, 3919, 122, 16668, 51499], "temperature": 0, "avg_logprob": -0.09122826128589864, "compression_ratio": 1.3231707317073171, "no_speech_prob": 1.4075023264947806e-11}, {"id": 53, "seek": 8520, "start": 107.88, "end": 109.58, "text": "你必须透过又花时间", "tokens": [51499, 2166, 28531, 10178, 119, 45808, 16866, 17047, 20127, 44848, 51584], "temperature": 0, "avg_logprob": -0.09122826128589864, "compression_ratio": 1.3231707317073171, "no_speech_prob": 1.4075023264947806e-11}, {"id": 54, "seek": 8520, "start": 109.58, "end": 111.58, "text": "又花钱的光学资源辨识", "tokens": [51584, 17047, 20127, 39623, 1546, 20690, 29618, 5266, 226, 47402, 9830, 101, 5233, 228, 51684], "temperature": 0, "avg_logprob": -0.09122826128589864, "compression_ratio": 1.3231707317073171, "no_speech_prob": 1.4075023264947806e-11}, {"id": 55, "seek": 8520, "start": 111.58, "end": 112.56, "text": "也就是OCR", "tokens": [51684, 6404, 5620, 46, 18547, 51733], "temperature": 0, "avg_logprob": -0.09122826128589864, "compression_ratio": 1.3231707317073171, "no_speech_prob": 1.4075023264947806e-11}, {"id": 56, "seek": 11256, "start": 112.56, "end": 114.24000000000001, "text": "才有办法把文字读出来", "tokens": [50365, 18888, 2412, 39453, 11148, 16075, 17174, 22381, 5233, 119, 44561, 50449], "temperature": 0, "avg_logprob": -0.0987328569939796, "compression_ratio": 1.2550724637681159, "no_speech_prob": 1.1987371165145255e-11}, {"id": 57, "seek": 11256, "start": 114.24000000000001, "end": 115.88, "text": "那传统的做法是怎么样呢", "tokens": [50449, 4184, 7384, 254, 10115, 253, 1546, 10907, 11148, 1541, 48200, 6240, 50531], "temperature": 0, "avg_logprob": -0.0987328569939796, "compression_ratio": 1.2550724637681159, "no_speech_prob": 1.1987371165145255e-11}, {"id": 58, "seek": 11256, "start": 115.88, "end": 116.96000000000001, "text": "为了保险起见", "tokens": [50531, 13992, 2289, 24302, 8842, 102, 9147, 23813, 50585], "temperature": 0, "avg_logprob": -0.0987328569939796, "compression_ratio": 1.2550724637681159, "no_speech_prob": 1.1987371165145255e-11}, {"id": 59, "seek": 11256, "start": 116.96000000000001, "end": 119.02, "text": "大家常常不管三七二十一", "tokens": [50585, 6868, 11279, 11279, 45376, 10960, 29920, 11217, 20145, 2257, 50688], "temperature": 0, "avg_logprob": -0.0987328569939796, "compression_ratio": 1.2550724637681159, "no_speech_prob": 1.1987371165145255e-11}, {"id": 60, "seek": 11256, "start": 119.02, "end": 121.86, "text": "把所有PDF全部丢给OCR服务处理", "tokens": [50688, 16075, 39300, 17349, 37, 38714, 940, 95, 23197, 46, 18547, 27408, 5087, 94, 1787, 226, 13876, 50830], "temperature": 0, "avg_logprob": -0.0987328569939796, "compression_ratio": 1.2550724637681159, "no_speech_prob": 1.1987371165145255e-11}, {"id": 61, "seek": 11256, "start": 121.86, "end": 123.48, "text": "这种一刀切的做法", "tokens": [50830, 5562, 39810, 2257, 2437, 222, 23632, 1546, 10907, 11148, 50911], "temperature": 0, "avg_logprob": -0.0987328569939796, "compression_ratio": 1.2550724637681159, "no_speech_prob": 1.1987371165145255e-11}, {"id": 62, "seek": 11256, "start": 123.48, "end": 125.46000000000001, "text": "不只让云端服务的费用暴增", "tokens": [50911, 1960, 14003, 33650, 1369, 239, 11957, 107, 27408, 5087, 94, 1546, 18464, 117, 9254, 23826, 112, 24228, 252, 51010], "temperature": 0, "avg_logprob": -0.0987328569939796, "compression_ratio": 1.2550724637681159, "no_speech_prob": 1.1987371165145255e-11}, {"id": 63, "seek": 11256, "start": 125.46000000000001, "end": 127.34, "text": "也造成了严重的处理延迟", "tokens": [51010, 6404, 37583, 11336, 2289, 940, 98, 12624, 1546, 1787, 226, 13876, 17527, 114, 3316, 253, 51104], "temperature": 0, "avg_logprob": -0.0987328569939796, "compression_ratio": 1.2550724637681159, "no_speech_prob": 1.1987371165145255e-11}, {"id": 64, "seek": 11256, "start": 127.34, "end": 129.52, "text": "而PDF Inspector这个专案", "tokens": [51104, 11070, 17349, 37, 33402, 15368, 940, 241, 28899, 51213], "temperature": 0, "avg_logprob": -0.0987328569939796, "compression_ratio": 1.2550724637681159, "no_speech_prob": 1.1987371165145255e-11}, {"id": 65, "seek": 11256, "start": 129.52, "end": 131.28, "text": "切入的正是这个痛点", "tokens": [51213, 23632, 14028, 1546, 15789, 1541, 15368, 28737, 12579, 51301], "temperature": 0, "avg_logprob": -0.0987328569939796, "compression_ratio": 1.2550724637681159, "no_speech_prob": 1.1987371165145255e-11}, {"id": 66, "seek": 11256, "start": 131.28, "end": 134.04, "text": "它不做那种什么都包的大杂烩功能", "tokens": [51301, 11284, 1960, 10907, 4184, 39810, 10440, 7182, 23305, 1546, 3582, 4422, 224, 23661, 102, 32311, 8225, 51439], "temperature": 0, "avg_logprob": -0.0987328569939796, "compression_ratio": 1.2550724637681159, "no_speech_prob": 1.1987371165145255e-11}, {"id": 67, "seek": 11256, "start": 134.04, "end": 135.56, "text": "而是专心做好第一步", "tokens": [51439, 11070, 1541, 940, 241, 7945, 10907, 2131, 18049, 31429, 51515], "temperature": 0, "avg_logprob": -0.0987328569939796, "compression_ratio": 1.2550724637681159, "no_speech_prob": 1.1987371165145255e-11}, {"id": 68, "seek": 11256, "start": 135.56, "end": 137.58, "text": "用最快的速度判断出", "tokens": [51515, 9254, 8661, 10251, 1546, 31217, 13127, 2437, 97, 4307, 255, 7781, 51616], "temperature": 0, "avg_logprob": -0.0987328569939796, "compression_ratio": 1.2550724637681159, "no_speech_prob": 1.1987371165145255e-11}, {"id": 69, "seek": 11256, "start": 137.58, "end": 140.24, "text": "这份PDF到底需不需要跑OCR", "tokens": [51616, 5562, 36266, 17349, 37, 33883, 32535, 1960, 35748, 32585, 46, 18547, 51749], "temperature": 0, "avg_logprob": -0.0987328569939796, "compression_ratio": 1.2550724637681159, "no_speech_prob": 1.1987371165145255e-11}, {"id": 70, "seek": 14024, "start": 140.24, "end": 143.10000000000002, "text": "PDF Inspector之所以这么有效率", "tokens": [50365, 17349, 37, 33402, 9574, 7239, 31889, 2412, 43076, 44866, 50508], "temperature": 0, "avg_logprob": -0.08073711395263672, "compression_ratio": 1.1754966887417218, "no_speech_prob": 1.5261987854064962e-11}, {"id": 71, "seek": 14024, "start": 143.10000000000002, "end": 145.76000000000002, "text": "关键就在它精巧的架构设计", "tokens": [50508, 28053, 23049, 106, 3111, 3581, 11284, 34910, 43605, 1546, 7360, 114, 7360, 226, 7422, 122, 7422, 94, 50641], "temperature": 0, "avg_logprob": -0.08073711395263672, "compression_ratio": 1.1754966887417218, "no_speech_prob": 1.5261987854064962e-11}, {"id": 72, "seek": 14024, "start": 145.76000000000002, "end": 147.02, "text": "你可以把它想象成", "tokens": [50641, 42766, 42061, 7093, 45007, 11336, 50704], "temperature": 0, "avg_logprob": -0.08073711395263672, "compression_ratio": 1.1754966887417218, "no_speech_prob": 1.5261987854064962e-11}, {"id": 73, "seek": 14024, "start": 147.02, "end": 148.88, "text": "医院的减伤分类站", "tokens": [50704, 9937, 119, 38358, 1546, 6336, 237, 7384, 97, 6627, 22113, 119, 34155, 50797], "temperature": 0, "avg_logprob": -0.08073711395263672, "compression_ratio": 1.1754966887417218, "no_speech_prob": 1.5261987854064962e-11}, {"id": 74, "seek": 14024, "start": 148.88, "end": 150.8, "text": "当一份PDF文件送进来", "tokens": [50797, 16233, 2257, 36266, 17349, 37, 17174, 20485, 29309, 36700, 6912, 50893], "temperature": 0, "avg_logprob": -0.08073711395263672, "compression_ratio": 1.1754966887417218, "no_speech_prob": 1.5261987854064962e-11}, {"id": 75, "seek": 14024, "start": 150.8, "end": 153.56, "text": "它不会马上做全身健康检查", "tokens": [50893, 11284, 47928, 49970, 5708, 10907, 11319, 19847, 44201, 6346, 115, 16407, 222, 42623, 51031], "temperature": 0, "avg_logprob": -0.08073711395263672, "compression_ratio": 1.1754966887417218, "no_speech_prob": 1.5261987854064962e-11}, {"id": 76, "seek": 14024, "start": 153.56, "end": 155.20000000000002, "text": "相反地会先交给一个", "tokens": [51031, 15106, 22138, 10928, 12949, 10108, 28455, 23197, 20182, 51113], "temperature": 0, "avg_logprob": -0.08073711395263672, "compression_ratio": 1.1754966887417218, "no_speech_prob": 1.5261987854064962e-11}, {"id": 77, "seek": 14024, "start": 155.20000000000002, "end": 157.32000000000002, "text": "速度超快的侦测器模组", "tokens": [51113, 31217, 13127, 19869, 10251, 1546, 3254, 99, 11038, 233, 34386, 41908, 10115, 226, 51219], "temperature": 0, "avg_logprob": -0.08073711395263672, "compression_ratio": 1.1754966887417218, "no_speech_prob": 1.5261987854064962e-11}, {"id": 78, "seek": 14024, "start": 157.32000000000002, "end": 159.18, "text": "这个侦测器很聪明", "tokens": [51219, 15368, 3254, 99, 11038, 233, 34386, 4563, 8171, 103, 11100, 51312], "temperature": 0, "avg_logprob": -0.08073711395263672, "compression_ratio": 1.1754966887417218, "no_speech_prob": 1.5261987854064962e-11}, {"id": 79, "seek": 14024, "start": 159.18, "end": 160.52, "text": "它不读完整个档案", "tokens": [51312, 11284, 1960, 5233, 119, 14128, 27662, 7549, 21416, 96, 28899, 51379], "temperature": 0, "avg_logprob": -0.08073711395263672, "compression_ratio": 1.1754966887417218, "no_speech_prob": 1.5261987854064962e-11}, {"id": 80, "seek": 14024, "start": 160.52, "end": 162.94, "text": "只抽样检查PDF里面", "tokens": [51379, 14003, 46022, 14496, 16407, 222, 42623, 17349, 37, 15759, 8833, 51500], "temperature": 0, "avg_logprob": -0.08073711395263672, "compression_ratio": 1.1754966887417218, "no_speech_prob": 1.5261987854064962e-11}, {"id": 81, "seek": 14024, "start": 162.94, "end": 165.38, "text": "那些代表文字或图像的操作指令", "tokens": [51500, 4184, 13824, 19105, 17571, 17174, 22381, 19780, 3919, 122, 12760, 1546, 13167, 235, 11914, 25922, 49061, 51622], "temperature": 0, "avg_logprob": -0.08073711395263672, "compression_ratio": 1.1754966887417218, "no_speech_prob": 1.5261987854064962e-11}, {"id": 82, "seek": 16538, "start": 165.38, "end": 167.4, "text": "光是靠着分析这些指令", "tokens": [50365, 20690, 1541, 5363, 254, 20708, 6627, 7360, 238, 5562, 13824, 25922, 49061, 50466], "temperature": 0, "avg_logprob": -0.0716108986027333, "compression_ratio": 1.3201320132013201, "no_speech_prob": 1.3100685467004602e-11}, {"id": 83, "seek": 16538, "start": 167.4, "end": 169.92, "text": "它就能在短短10到50毫秒内", "tokens": [50466, 11284, 3111, 8225, 3581, 5881, 255, 5881, 255, 3279, 4511, 2803, 7256, 104, 20956, 34742, 50592], "temperature": 0, "avg_logprob": -0.0716108986027333, "compression_ratio": 1.3201320132013201, "no_speech_prob": 1.3100685467004602e-11}, {"id": 84, "seek": 16538, "start": 169.92, "end": 172.7, "text": "判断出这份文件是文字型", "tokens": [50592, 2437, 97, 4307, 255, 7781, 5562, 36266, 17174, 20485, 1541, 17174, 22381, 39823, 50731], "temperature": 0, "avg_logprob": -0.0716108986027333, "compression_ratio": 1.3201320132013201, "no_speech_prob": 1.3100685467004602e-11}, {"id": 85, "seek": 16538, "start": 172.7, "end": 174.78, "text": "扫描型还是混合型", "tokens": [50731, 3416, 104, 11673, 237, 39823, 45726, 48640, 14245, 39823, 50835], "temperature": 0, "avg_logprob": -0.0716108986027333, "compression_ratio": 1.3201320132013201, "no_speech_prob": 1.3100685467004602e-11}, {"id": 86, "seek": 16538, "start": 174.78, "end": 177.94, "text": "这个速度就是它甩开其他工具的关键", "tokens": [50835, 15368, 31217, 13127, 5620, 11284, 3526, 102, 18937, 9572, 5000, 23323, 39806, 1546, 28053, 23049, 106, 50993], "temperature": 0, "avg_logprob": -0.0716108986027333, "compression_ratio": 1.3201320132013201, "no_speech_prob": 1.3100685467004602e-11}, {"id": 87, "seek": 16538, "start": 177.94, "end": 179.56, "text": "如果判断是文字型", "tokens": [50993, 13119, 2437, 97, 4307, 255, 1541, 17174, 22381, 39823, 51074], "temperature": 0, "avg_logprob": -0.0716108986027333, "compression_ratio": 1.3201320132013201, "no_speech_prob": 1.3100685467004602e-11}, {"id": 88, "seek": 16538, "start": 179.56, "end": 182.24, "text": "档案才会被交给下一步的提取器", "tokens": [51074, 21416, 96, 28899, 18888, 12949, 23238, 28455, 23197, 45271, 31429, 1546, 20949, 29436, 34386, 51208], "temperature": 0, "avg_logprob": -0.0716108986027333, "compression_ratio": 1.3201320132013201, "no_speech_prob": 1.3100685467004602e-11}, {"id": 89, "seek": 16538, "start": 182.24, "end": 184.35999999999999, "text": "模组来做深度的处理", "tokens": [51208, 41908, 10115, 226, 6912, 10907, 24043, 13127, 1546, 1787, 226, 13876, 51314], "temperature": 0, "avg_logprob": -0.0716108986027333, "compression_ratio": 1.3201320132013201, "no_speech_prob": 1.3100685467004602e-11}, {"id": 90, "seek": 16538, "start": 184.35999999999999, "end": 186.06, "text": "整个流程最精彩的设计", "tokens": [51314, 27662, 7549, 27854, 29649, 8661, 34910, 7391, 102, 1546, 7422, 122, 7422, 94, 51399], "temperature": 0, "avg_logprob": -0.0716108986027333, "compression_ratio": 1.3201320132013201, "no_speech_prob": 1.3100685467004602e-11}, {"id": 91, "seek": 16538, "start": 186.06, "end": 187.5, "text": "就是PDF档案", "tokens": [51399, 5620, 17349, 37, 21416, 96, 28899, 51471], "temperature": 0, "avg_logprob": -0.0716108986027333, "compression_ratio": 1.3201320132013201, "no_speech_prob": 1.3100685467004602e-11}, {"id": 92, "seek": 16538, "start": 187.5, "end": 189.35999999999999, "text": "从头到尾只会被读取一次", "tokens": [51471, 35630, 39862, 4511, 1530, 122, 14003, 12949, 23238, 5233, 119, 29436, 27505, 51564], "temperature": 0, "avg_logprob": -0.0716108986027333, "compression_ratio": 1.3201320132013201, "no_speech_prob": 1.3100685467004602e-11}, {"id": 93, "seek": 16538, "start": 189.35999999999999, "end": 191.24, "text": "它的分析结果会由", "tokens": [51564, 45224, 6627, 7360, 238, 45641, 9319, 12949, 23786, 51658], "temperature": 0, "avg_logprob": -0.0716108986027333, "compression_ratio": 1.3201320132013201, "no_speech_prob": 1.3100685467004602e-11}, {"id": 94, "seek": 16538, "start": 191.24, "end": 193.68, "text": "侦测和提取这两个模组共享", "tokens": [51658, 3254, 99, 11038, 233, 12565, 20949, 29436, 5562, 36257, 7549, 41908, 10115, 226, 15408, 27654, 51780], "temperature": 0, "avg_logprob": -0.0716108986027333, "compression_ratio": 1.3201320132013201, "no_speech_prob": 1.3100685467004602e-11}, {"id": 95, "seek": 19368, "start": 193.68, "end": 196.58, "text": "这样就彻底避免了重复读取和运算的浪费", "tokens": [50365, 21209, 3111, 7391, 119, 22816, 3330, 123, 2347, 235, 2289, 12624, 1787, 235, 5233, 119, 29436, 12565, 3316, 238, 19497, 1546, 11038, 103, 18464, 117, 50510], "temperature": 0, "avg_logprob": -0.07470426754075654, "compression_ratio": 1.2453333333333334, "no_speech_prob": 1.2665759933916387e-11}, {"id": 96, "seek": 19368, "start": 196.58, "end": 199.12, "text": "这种单次载入多方共享的架构", "tokens": [50510, 5562, 39810, 47446, 9487, 17819, 121, 14028, 6392, 9249, 15408, 27654, 1546, 7360, 114, 7360, 226, 50637], "temperature": 0, "avg_logprob": -0.07470426754075654, "compression_ratio": 1.2453333333333334, "no_speech_prob": 1.2665759933916387e-11}, {"id": 97, "seek": 19368, "start": 199.12, "end": 201.54000000000002, "text": "正是它能够又快又准的核心秘密", "tokens": [50637, 15789, 1541, 11284, 8225, 1787, 253, 17047, 10251, 17047, 6336, 228, 1546, 8987, 116, 7945, 8204, 246, 47838, 50758], "temperature": 0, "avg_logprob": -0.07470426754075654, "compression_ratio": 1.2453333333333334, "no_speech_prob": 1.2665759933916387e-11}, {"id": 98, "seek": 19368, "start": 201.54000000000002, "end": 202.98000000000002, "text": "那在实际应用上", "tokens": [50758, 4184, 3581, 24726, 8842, 227, 44297, 9254, 5708, 50830], "temperature": 0, "avg_logprob": -0.07470426754075654, "compression_ratio": 1.2453333333333334, "no_speech_prob": 1.2665759933916387e-11}, {"id": 99, "seek": 19368, "start": 202.98000000000002, "end": 205.58, "text": "PDF Inspector的场景就非常清楚了", "tokens": [50830, 17349, 37, 33402, 1546, 50255, 50218, 3111, 14392, 38433, 2289, 50960], "temperature": 0, "avg_logprob": -0.07470426754075654, "compression_ratio": 1.2453333333333334, "no_speech_prob": 1.2665759933916387e-11}, {"id": 100, "seek": 19368, "start": 205.58, "end": 206.02, "text": "首先", "tokens": [50960, 36490, 50982], "temperature": 0, "avg_logprob": -0.07470426754075654, "compression_ratio": 1.2453333333333334, "no_speech_prob": 1.2665759933916387e-11}, {"id": 101, "seek": 19368, "start": 206.02, "end": 208.54000000000002, "text": "想象一下那些需要处理大量报告", "tokens": [50982, 7093, 45007, 8861, 4184, 13824, 35748, 1787, 226, 13876, 3582, 26748, 49817, 16846, 51108], "temperature": 0, "avg_logprob": -0.07470426754075654, "compression_ratio": 1.2453333333333334, "no_speech_prob": 1.2665759933916387e-11}, {"id": 102, "seek": 19368, "start": 208.54000000000002, "end": 211.42000000000002, "text": "发票或法律文件的后端系统", "tokens": [51108, 28926, 41925, 19780, 11148, 2172, 233, 17174, 20485, 1546, 13547, 11957, 107, 25368, 10115, 253, 51252], "temperature": 0, "avg_logprob": -0.07470426754075654, "compression_ratio": 1.2453333333333334, "no_speech_prob": 1.2665759933916387e-11}, {"id": 103, "seek": 19368, "start": 211.42000000000002, "end": 214.4, "text": "开发者可以把PDF Inspector当作整个流程的第一关", "tokens": [51252, 18937, 28926, 12444, 6723, 16075, 17349, 37, 33402, 16233, 11914, 27662, 7549, 27854, 29649, 1546, 18049, 28053, 51401], "temperature": 0, "avg_logprob": -0.07470426754075654, "compression_ratio": 1.2453333333333334, "no_speech_prob": 1.2665759933916387e-11}, {"id": 104, "seek": 19368, "start": 214.4, "end": 216.24, "text": "它就像一个智慧路由器", "tokens": [51401, 11284, 3111, 12760, 20182, 5094, 118, 12358, 100, 24658, 23786, 34386, 51493], "temperature": 0, "avg_logprob": -0.07470426754075654, "compression_ratio": 1.2453333333333334, "no_speech_prob": 1.2665759933916387e-11}, {"id": 105, "seek": 19368, "start": 216.24, "end": 218.26000000000002, "text": "可以快速的把文字型PDF", "tokens": [51493, 6723, 10251, 31217, 1546, 16075, 17174, 22381, 39823, 17349, 37, 51594], "temperature": 0, "avg_logprob": -0.07470426754075654, "compression_ratio": 1.2453333333333334, "no_speech_prob": 1.2665759933916387e-11}, {"id": 106, "seek": 19368, "start": 218.26000000000002, "end": 220.22, "text": "指向本地的服务来快速提取", "tokens": [51594, 25922, 24282, 8802, 10928, 1546, 27408, 5087, 94, 6912, 10251, 31217, 20949, 29436, 51692], "temperature": 0, "avg_logprob": -0.07470426754075654, "compression_ratio": 1.2453333333333334, "no_speech_prob": 1.2665759933916387e-11}, {"id": 107, "seek": 19368, "start": 220.22, "end": 221.94, "text": "只有那些真正的扫描件", "tokens": [51692, 35244, 4184, 13824, 6303, 15789, 1546, 3416, 104, 11673, 237, 20485, 51778], "temperature": 0, "avg_logprob": -0.07470426754075654, "compression_ratio": 1.2453333333333334, "no_speech_prob": 1.2665759933916387e-11}, {"id": 108, "seek": 22194, "start": 221.94, "end": 224.6, "text": "才需要送去昂贵的云端OCR", "tokens": [50365, 18888, 35748, 29309, 6734, 1431, 224, 18464, 113, 1546, 1369, 239, 11957, 107, 46, 18547, 50498], "temperature": 0, "avg_logprob": -0.10814127107945884, "compression_ratio": 1.1968911917098446, "no_speech_prob": 1.4068587440851932e-11}, {"id": 109, "seek": 22194, "start": 224.6, "end": 227.18, "text": "这样一来成本和延迟都能大幅降低", "tokens": [50498, 21209, 2257, 6912, 11336, 8802, 12565, 17527, 114, 3316, 253, 7182, 8225, 3582, 3509, 227, 47421, 41377, 50627], "temperature": 0, "avg_logprob": -0.10814127107945884, "compression_ratio": 1.1968911917098446, "no_speech_prob": 1.4068587440851932e-11}, {"id": 110, "seek": 22194, "start": 227.18, "end": 229.2, "text": "再来在AI HAN REC", "tokens": [50627, 8623, 6912, 3581, 48698, 389, 1770, 497, 8140, 50728], "temperature": 0, "avg_logprob": -0.10814127107945884, "compression_ratio": 1.1968911917098446, "no_speech_prob": 1.4068587440851932e-11}, {"id": 111, "seek": 22194, "start": 229.2, "end": 231.62, "text": "也就是检索增强生成的应用里", "tokens": [50728, 6404, 5620, 16407, 222, 7732, 95, 24228, 252, 5702, 118, 8244, 11336, 1546, 44297, 9254, 15759, 50849], "temperature": 0, "avg_logprob": -0.10814127107945884, "compression_ratio": 1.1968911917098446, "no_speech_prob": 1.4068587440851932e-11}, {"id": 112, "seek": 22194, "start": 231.62, "end": 233.48, "text": "数据科学家可以用它", "tokens": [50849, 33188, 26075, 106, 42091, 29618, 5155, 6723, 9254, 11284, 50942], "temperature": 0, "avg_logprob": -0.10814127107945884, "compression_ratio": 1.1968911917098446, "no_speech_prob": 1.4068587440851932e-11}, {"id": 113, "seek": 22194, "start": 233.48, "end": 235.9, "text": "快速把研究论文这类的PDF", "tokens": [50942, 10251, 31217, 16075, 23230, 242, 44704, 7422, 118, 17174, 5562, 22113, 119, 1546, 17349, 37, 51063], "temperature": 0, "avg_logprob": -0.10814127107945884, "compression_ratio": 1.1968911917098446, "no_speech_prob": 1.4068587440851932e-11}, {"id": 114, "seek": 22194, "start": 235.9, "end": 238.18, "text": "转成结构清楚的Markdown格式", "tokens": [51063, 17819, 105, 11336, 45641, 7360, 226, 38433, 1546, 15168, 5093, 30921, 27584, 51177], "temperature": 0, "avg_logprob": -0.10814127107945884, "compression_ratio": 1.1968911917098446, "no_speech_prob": 1.4068587440851932e-11}, {"id": 115, "seek": 22194, "start": 238.18, "end": 240.7, "text": "直接拿来当作大型语言模型的训练资料", "tokens": [51177, 43297, 24351, 6912, 16233, 11914, 3582, 39823, 5233, 255, 12009, 41908, 39823, 1546, 7422, 255, 10115, 225, 5266, 226, 33404, 51303], "temperature": 0, "avg_logprob": -0.10814127107945884, "compression_ratio": 1.1968911917098446, "no_speech_prob": 1.4068587440851932e-11}, {"id": 116, "seek": 22194, "start": 240.7, "end": 241.5, "text": "更酷的是", "tokens": [51303, 19002, 16911, 115, 24620, 51343], "temperature": 0, "avg_logprob": -0.10814127107945884, "compression_ratio": 1.1968911917098446, "no_speech_prob": 1.4068587440851932e-11}, {"id": 117, "seek": 22194, "start": 241.5, "end": 243.7, "text": "它还提供了Web Assembly版本", "tokens": [51343, 11284, 14852, 20949, 3254, 249, 2289, 4360, 65, 20399, 42096, 8802, 51453], "temperature": 0, "avg_logprob": -0.10814127107945884, "compression_ratio": 1.1968911917098446, "no_speech_prob": 1.4068587440851932e-11}, {"id": 118, "seek": 22194, "start": 243.7, "end": 244.4, "text": "这代表什么", "tokens": [51453, 5562, 19105, 17571, 10440, 51488], "temperature": 0, "avg_logprob": -0.10814127107945884, "compression_ratio": 1.1968911917098446, "no_speech_prob": 1.4068587440851932e-11}, {"id": 119, "seek": 22194, "start": 244.4, "end": 247.02, "text": "这代表前端开发者可以直接在浏览器里", "tokens": [51488, 5562, 19105, 17571, 8945, 11957, 107, 18937, 28926, 12444, 6723, 43297, 3581, 11038, 237, 6758, 230, 34386, 15759, 51619], "temperature": 0, "avg_logprob": -0.10814127107945884, "compression_ratio": 1.1968911917098446, "no_speech_prob": 1.4068587440851932e-11}, {"id": 120, "seek": 22194, "start": 247.02, "end": 249.28, "text": "就分析使用者上传的PDF", "tokens": [51619, 3111, 6627, 7360, 238, 22982, 9254, 12444, 5708, 7384, 254, 1546, 17349, 37, 51732], "temperature": 0, "avg_logprob": -0.10814127107945884, "compression_ratio": 1.1968911917098446, "no_speech_prob": 1.4068587440851932e-11}, {"id": 121, "seek": 22194, "start": 249.28, "end": 251.82, "text": "完全不需要跟伺服器来回沟通", "tokens": [51732, 37100, 1960, 35748, 9678, 7384, 118, 27408, 34386, 6912, 8350, 3308, 253, 19550, 51859], "temperature": 0, "avg_logprob": -0.10814127107945884, "compression_ratio": 1.1968911917098446, "no_speech_prob": 1.4068587440851932e-11}, {"id": 122, "seek": 25182, "start": 251.82, "end": 253.34, "text": "既能提升效能", "tokens": [50365, 4479, 95, 8225, 20949, 41670, 43076, 8225, 50441], "temperature": 0, "avg_logprob": -0.08828800827709597, "compression_ratio": 1.195187165775401, "no_speech_prob": 1.2128915059805845e-11}, {"id": 123, "seek": 25182, "start": 253.34, "end": 254.92, "text": "又能保护使用者隐私", "tokens": [50441, 17047, 8225, 24302, 7235, 97, 22982, 9254, 12444, 10673, 238, 20083, 50520], "temperature": 0, "avg_logprob": -0.08828800827709597, "compression_ratio": 1.195187165775401, "no_speech_prob": 1.2128915059805845e-11}, {"id": 124, "seek": 25182, "start": 254.92, "end": 257.64, "text": "而且因为专案提供了Python Node.js", "tokens": [50520, 22942, 34627, 940, 241, 28899, 20949, 3254, 249, 2289, 47, 88, 11943, 38640, 13, 25530, 50656], "temperature": 0, "avg_logprob": -0.08828800827709597, "compression_ratio": 1.195187165775401, "no_speech_prob": 1.2128915059805845e-11}, {"id": 125, "seek": 25182, "start": 257.64, "end": 258.98, "text": "这些主流语言的绑定", "tokens": [50656, 5562, 13824, 13557, 27854, 5233, 255, 12009, 1546, 10115, 239, 12088, 50723], "temperature": 0, "avg_logprob": -0.08828800827709597, "compression_ratio": 1.195187165775401, "no_speech_prob": 1.2128915059805845e-11}, {"id": 126, "seek": 25182, "start": 258.98, "end": 260.62, "text": "上手的门槛也相当低", "tokens": [50723, 5708, 11389, 1546, 8259, 101, 24880, 249, 6404, 15106, 16233, 41377, 50805], "temperature": 0, "avg_logprob": -0.08828800827709597, "compression_ratio": 1.195187165775401, "no_speech_prob": 1.2128915059805845e-11}, {"id": 127, "seek": 25182, "start": 260.62, "end": 263.15999999999997, "text": "社群对PDF Inspector的反应", "tokens": [50805, 27658, 7248, 97, 8713, 17349, 37, 33402, 1546, 22138, 44297, 50932], "temperature": 0, "avg_logprob": -0.08828800827709597, "compression_ratio": 1.195187165775401, "no_speech_prob": 1.2128915059805845e-11}, {"id": 128, "seek": 25182, "start": 263.15999999999997, "end": 264.32, "text": "普遍都非常好", "tokens": [50932, 29993, 3330, 235, 7182, 14392, 2131, 50990], "temperature": 0, "avg_logprob": -0.08828800827709597, "compression_ratio": 1.195187165775401, "no_speech_prob": 1.2128915059805845e-11}, {"id": 129, "seek": 25182, "start": 264.32, "end": 265.38, "text": "大家称赞的点", "tokens": [50990, 6868, 8204, 108, 5266, 252, 1546, 12579, 51043], "temperature": 0, "avg_logprob": -0.08828800827709597, "compression_ratio": 1.195187165775401, "no_speech_prob": 1.2128915059805845e-11}, {"id": 130, "seek": 25182, "start": 265.38, "end": 267.08, "text": "主要都集中在它那", "tokens": [51043, 13557, 4275, 7182, 26020, 5975, 3581, 11284, 4184, 51128], "temperature": 0, "avg_logprob": -0.08828800827709597, "compression_ratio": 1.195187165775401, "no_speech_prob": 1.2128915059805845e-11}, {"id": 131, "seek": 25182, "start": 267.08, "end": 268.9, "text": "快到下人的处理速度", "tokens": [51128, 10251, 4511, 4438, 4035, 1546, 1787, 226, 13876, 31217, 13127, 51219], "temperature": 0, "avg_logprob": -0.08828800827709597, "compression_ratio": 1.195187165775401, "no_speech_prob": 1.2128915059805845e-11}, {"id": 132, "seek": 25182, "start": 268.9, "end": 271.6, "text": "还有完全可以在本地端执行的能力", "tokens": [51219, 35091, 37100, 6723, 3581, 8802, 10928, 11957, 107, 3416, 100, 8082, 1546, 8225, 13486, 51354], "temperature": 0, "avg_logprob": -0.08828800827709597, "compression_ratio": 1.195187165775401, "no_speech_prob": 1.2128915059805845e-11}, {"id": 133, "seek": 25182, "start": 271.6, "end": 272.78, "text": "很多开发者都说", "tokens": [51354, 20778, 18937, 28926, 12444, 7182, 8090, 51413], "temperature": 0, "avg_logprob": -0.08828800827709597, "compression_ratio": 1.195187165775401, "no_speech_prob": 1.2128915059805845e-11}, {"id": 134, "seek": 25182, "start": 272.78, "end": 274.08, "text": "光是它提供的", "tokens": [51413, 20690, 1541, 11284, 20949, 3254, 249, 1546, 51478], "temperature": 0, "avg_logprob": -0.08828800827709597, "compression_ratio": 1.195187165775401, "no_speech_prob": 1.2128915059805845e-11}, {"id": 135, "seek": 25182, "start": 274.08, "end": 276.21999999999997, "text": "需要OCR的页面列表这个功能", "tokens": [51478, 35748, 46, 18547, 1546, 10178, 113, 8833, 43338, 17571, 15368, 32311, 8225, 51585], "temperature": 0, "avg_logprob": -0.08828800827709597, "compression_ratio": 1.195187165775401, "no_speech_prob": 1.2128915059805845e-11}, {"id": 136, "seek": 25182, "start": 276.21999999999997, "end": 277.36, "text": "就帮他们省下了", "tokens": [51585, 3111, 4845, 106, 47911, 2862, 223, 4438, 2289, 51642], "temperature": 0, "avg_logprob": -0.08828800827709597, "compression_ratio": 1.195187165775401, "no_speech_prob": 1.2128915059805845e-11}, {"id": 137, "seek": 25182, "start": 277.36, "end": 278.71999999999997, "text": "大笔的云端服务账单", "tokens": [51642, 3582, 5437, 242, 1546, 1369, 239, 11957, 107, 27408, 5087, 94, 18464, 99, 47446, 51710], "temperature": 0, "avg_logprob": -0.08828800827709597, "compression_ratio": 1.195187165775401, "no_speech_prob": 1.2128915059805845e-11}, {"id": 138, "seek": 25182, "start": 278.71999999999997, "end": 280.03999999999996, "text": "社群也特别提到", "tokens": [51710, 27658, 7248, 97, 6404, 17682, 18453, 20949, 4511, 51776], "temperature": 0, "avg_logprob": -0.08828800827709597, "compression_ratio": 1.195187165775401, "no_speech_prob": 1.2128915059805845e-11}, {"id": 139, "seek": 28004, "start": 280.04, "end": 282.44, "text": "它输出的Markdown品质很好", "tokens": [50365, 11284, 9830, 241, 7781, 1546, 15168, 5093, 30246, 18464, 101, 23801, 50485], "temperature": 0, "avg_logprob": -0.08864830017089843, "compression_ratio": 1.1915492957746479, "no_speech_prob": 1.2467376089841942e-11}, {"id": 140, "seek": 28004, "start": 282.44, "end": 285.64000000000004, "text": "非常适合直接喂给大型语言模型", "tokens": [50485, 14392, 2215, 224, 14245, 43297, 35142, 23197, 3582, 39823, 5233, 255, 12009, 41908, 39823, 50645], "temperature": 0, "avg_logprob": -0.08864830017089843, "compression_ratio": 1.1915492957746479, "no_speech_prob": 1.2467376089841942e-11}, {"id": 141, "seek": 28004, "start": 285.64000000000004, "end": 287.70000000000005, "text": "目前还没有看到什么明显的负评", "tokens": [50645, 39004, 14852, 17944, 18032, 10440, 11100, 1431, 122, 1546, 18464, 253, 5233, 226, 50748], "temperature": 0, "avg_logprob": -0.08864830017089843, "compression_ratio": 1.1915492957746479, "no_speech_prob": 1.2467376089841942e-11}, {"id": 142, "seek": 28004, "start": 287.70000000000005, "end": 289.78000000000003, "text": "不过倒是有一些很务实的提醒", "tokens": [50748, 1960, 16866, 30878, 1541, 32241, 13824, 4563, 5087, 94, 24726, 1546, 20949, 40796, 50852], "temperature": 0, "avg_logprob": -0.08864830017089843, "compression_ratio": 1.1915492957746479, "no_speech_prob": 1.2467376089841942e-11}, {"id": 143, "seek": 28004, "start": 289.78000000000003, "end": 290.54, "text": "比方说", "tokens": [50852, 11706, 9249, 8090, 50890], "temperature": 0, "avg_logprob": -0.08864830017089843, "compression_ratio": 1.1915492957746479, "no_speech_prob": 1.2467376089841942e-11}, {"id": 144, "seek": 28004, "start": 290.54, "end": 293.02000000000004, "text": "这个专案本身不包含OCR功能", "tokens": [50890, 15368, 940, 241, 28899, 8802, 19847, 1960, 23305, 2392, 104, 46, 18547, 32311, 8225, 51014], "temperature": 0, "avg_logprob": -0.08864830017089843, "compression_ratio": 1.1915492957746479, "no_speech_prob": 1.2467376089841942e-11}, {"id": 145, "seek": 28004, "start": 293.02000000000004, "end": 294.58000000000004, "text": "所以如果遇到扫描文件", "tokens": [51014, 7239, 13119, 3330, 229, 4511, 3416, 104, 11673, 237, 17174, 20485, 51092], "temperature": 0, "avg_logprob": -0.08864830017089843, "compression_ratio": 1.1915492957746479, "no_speech_prob": 1.2467376089841942e-11}, {"id": 146, "seek": 28004, "start": 294.58000000000004, "end": 296.76, "text": "还是得搭配其他的引擎才能处理", "tokens": [51092, 45726, 5916, 22220, 255, 38846, 9572, 31309, 41301, 13167, 236, 18888, 8225, 1787, 226, 13876, 51201], "temperature": 0, "avg_logprob": -0.08864830017089843, "compression_ratio": 1.1915492957746479, "no_speech_prob": 1.2467376089841942e-11}, {"id": 147, "seek": 28004, "start": 296.76, "end": 298.84000000000003, "text": "如果我们把眼光放远一点来看", "tokens": [51201, 13119, 15003, 16075, 25281, 20690, 12744, 3316, 250, 49042, 6912, 4200, 51305], "temperature": 0, "avg_logprob": -0.08864830017089843, "compression_ratio": 1.1915492957746479, "no_speech_prob": 1.2467376089841942e-11}, {"id": 148, "seek": 28004, "start": 298.84000000000003, "end": 300.54, "text": "PDF Inspector的出现", "tokens": [51305, 17349, 37, 33402, 1546, 7781, 20204, 51390], "temperature": 0, "avg_logprob": -0.08864830017089843, "compression_ratio": 1.1915492957746479, "no_speech_prob": 1.2467376089841942e-11}, {"id": 149, "seek": 28004, "start": 300.54, "end": 303.84000000000003, "text": "其实正在挑战整个业界处理文件的预设思维", "tokens": [51390, 41646, 15789, 3581, 49067, 1486, 246, 27662, 7549, 940, 248, 21185, 1787, 226, 13876, 17174, 20485, 1546, 12501, 226, 7422, 122, 8870, 10115, 112, 51555], "temperature": 0, "avg_logprob": -0.08864830017089843, "compression_ratio": 1.1915492957746479, "no_speech_prob": 1.2467376089841942e-11}, {"id": 150, "seek": 28004, "start": 303.84000000000003, "end": 306.68, "text": "它所提倡的分类优先模式", "tokens": [51555, 11284, 5966, 20949, 2122, 94, 1546, 6627, 22113, 119, 7384, 246, 10108, 41908, 27584, 51697], "temperature": 0, "avg_logprob": -0.08864830017089843, "compression_ratio": 1.1915492957746479, "no_speech_prob": 1.2467376089841942e-11}, {"id": 151, "seek": 30668, "start": 306.68, "end": 311.14, "text": "会促使开发者从过去那种暴力破解式的OCR处理", "tokens": [50365, 12949, 7792, 225, 22982, 18937, 28926, 12444, 35630, 16866, 6734, 4184, 39810, 23826, 112, 13486, 39560, 17278, 27584, 1546, 46, 18547, 1787, 226, 13876, 50588], "temperature": 0, "avg_logprob": -0.12369751443668288, "compression_ratio": 1.1610486891385767, "no_speech_prob": 1.551450981573943e-11}, {"id": 152, "seek": 30668, "start": 311.14, "end": 312.38, "text": "转向更精细", "tokens": [50588, 17819, 105, 24282, 19002, 34910, 10115, 228, 50650], "temperature": 0, "avg_logprob": -0.12369751443668288, "compression_ratio": 1.1610486891385767, "no_speech_prob": 1.551450981573943e-11}, {"id": 153, "seek": 30668, "start": 312.38, "end": 314.94, "text": "也更有成本效益的智慧化流程", "tokens": [50650, 6404, 19002, 2412, 11336, 8802, 43076, 5419, 232, 1546, 5094, 118, 12358, 100, 23756, 27854, 29649, 50778], "temperature": 0, "avg_logprob": -0.12369751443668288, "compression_ratio": 1.1610486891385767, "no_speech_prob": 1.551450981573943e-11}, {"id": 154, "seek": 30668, "start": 314.94, "end": 315.48, "text": "不过", "tokens": [50778, 1960, 16866, 50805], "temperature": 0, "avg_logprob": -0.12369751443668288, "compression_ratio": 1.1610486891385767, "no_speech_prob": 1.551450981573943e-11}, {"id": 155, "seek": 30668, "start": 315.48, "end": 317.58, "text": "它的潜在风险也不能忽视", "tokens": [50805, 45224, 35622, 250, 3581, 47209, 8842, 102, 6404, 28590, 4263, 121, 40656, 50910], "temperature": 0, "avg_logprob": -0.12369751443668288, "compression_ratio": 1.1610486891385767, "no_speech_prob": 1.551450981573943e-11}, {"id": 156, "seek": 30668, "start": 317.58, "end": 318.14, "text": "首先", "tokens": [50910, 36490, 50938], "temperature": 0, "avg_logprob": -0.12369751443668288, "compression_ratio": 1.1610486891385767, "no_speech_prob": 1.551450981573943e-11}, {"id": 157, "seek": 30668, "start": 318.14, "end": 321.38, "text": "这个专案高度依赖底层的一个叫做", "tokens": [50938, 15368, 940, 241, 28899, 12979, 13127, 3254, 251, 5266, 244, 22816, 9636, 224, 1546, 20182, 19855, 10907, 51100], "temperature": 0, "avg_logprob": -0.12369751443668288, "compression_ratio": 1.1610486891385767, "no_speech_prob": 1.551450981573943e-11}, {"id": 158, "seek": 30668, "start": 321.38, "end": 323.08, "text": "LobDF的函事库", "tokens": [51100, 43, 996, 35, 37, 1546, 6336, 121, 6973, 6346, 241, 51185], "temperature": 0, "avg_logprob": -0.12369751443668288, "compression_ratio": 1.1610486891385767, "no_speech_prob": 1.551450981573943e-11}, {"id": 159, "seek": 30668, "start": 323.08, "end": 326.32, "text": "所以只要LobDF有任何限制", "tokens": [51185, 7239, 14003, 4275, 43, 996, 35, 37, 2412, 26443, 11906, 43446, 25491, 51347], "temperature": 0, "avg_logprob": -0.12369751443668288, "compression_ratio": 1.1610486891385767, "no_speech_prob": 1.551450981573943e-11}, {"id": 160, "seek": 30668, "start": 326.32, "end": 328.12, "text": "都会直接影响到它", "tokens": [51347, 7182, 12949, 43297, 16820, 3858, 235, 4511, 11284, 51437], "temperature": 0, "avg_logprob": -0.12369751443668288, "compression_ratio": 1.1610486891385767, "no_speech_prob": 1.551450981573943e-11}, {"id": 161, "seek": 30668, "start": 328.12, "end": 328.7, "text": "其次", "tokens": [51437, 9572, 9487, 51466], "temperature": 0, "avg_logprob": -0.12369751443668288, "compression_ratio": 1.1610486891385767, "no_speech_prob": 1.551450981573943e-11}, {"id": 162, "seek": 30668, "start": 328.7, "end": 330.56, "text": "它的分类和版面分析", "tokens": [51466, 45224, 6627, 22113, 119, 12565, 42096, 8833, 6627, 7360, 238, 51559], "temperature": 0, "avg_logprob": -0.12369751443668288, "compression_ratio": 1.1610486891385767, "no_speech_prob": 1.551450981573943e-11}, {"id": 163, "seek": 33056, "start": 330.56, "end": 333.02, "text": "大量采用了启发式演算法", "tokens": [50365, 3582, 26748, 5873, 229, 9254, 2289, 2392, 107, 28926, 27584, 31382, 19497, 11148, 50488], "temperature": 0, "avg_logprob": -0.09218816334390875, "compression_ratio": 1.239010989010989, "no_speech_prob": 1.1476121268561013e-11}, {"id": 164, "seek": 33056, "start": 333.02, "end": 333.84, "text": "这代表说", "tokens": [50488, 5562, 19105, 17571, 8090, 50529], "temperature": 0, "avg_logprob": -0.09218816334390875, "compression_ratio": 1.239010989010989, "no_speech_prob": 1.1476121268561013e-11}, {"id": 165, "seek": 33056, "start": 333.84, "end": 336.1, "text": "当它遇到那种特别复杂", "tokens": [50529, 16233, 11284, 3330, 229, 4511, 4184, 39810, 17682, 18453, 1787, 235, 4422, 224, 50642], "temperature": 0, "avg_logprob": -0.09218816334390875, "compression_ratio": 1.239010989010989, "no_speech_prob": 1.1476121268561013e-11}, {"id": 166, "seek": 33056, "start": 336.1, "end": 337.7, "text": "或是不照规矩来的版面时", "tokens": [50642, 19780, 1541, 1960, 32150, 6758, 226, 5881, 102, 6912, 1546, 42096, 8833, 15729, 50722], "temperature": 0, "avg_logprob": -0.09218816334390875, "compression_ratio": 1.239010989010989, "no_speech_prob": 1.1476121268561013e-11}, {"id": 167, "seek": 33056, "start": 337.7, "end": 339.46, "text": "还是有可能会判断错误", "tokens": [50722, 45726, 2412, 16657, 12949, 2437, 97, 4307, 255, 29900, 5233, 107, 50810], "temperature": 0, "avg_logprob": -0.09218816334390875, "compression_ratio": 1.239010989010989, "no_speech_prob": 1.1476121268561013e-11}, {"id": 168, "seek": 33056, "start": 339.46, "end": 339.94, "text": "最后", "tokens": [50810, 8661, 13547, 50834], "temperature": 0, "avg_logprob": -0.09218816334390875, "compression_ratio": 1.239010989010989, "no_speech_prob": 1.1476121268561013e-11}, {"id": 169, "seek": 33056, "start": 339.94, "end": 341.44, "text": "它清楚的功能边界", "tokens": [50834, 11284, 38433, 1546, 32311, 8225, 40503, 21185, 50909], "temperature": 0, "avg_logprob": -0.09218816334390875, "compression_ratio": 1.239010989010989, "no_speech_prob": 1.1476121268561013e-11}, {"id": 170, "seek": 33056, "start": 341.44, "end": 342.46, "text": "也代表使用者", "tokens": [50909, 6404, 19105, 17571, 22982, 9254, 12444, 50960], "temperature": 0, "avg_logprob": -0.09218816334390875, "compression_ratio": 1.239010989010989, "no_speech_prob": 1.1476121268561013e-11}, {"id": 171, "seek": 33056, "start": 342.46, "end": 344.9, "text": "必须自己去整合第三方的OCR服务", "tokens": [50960, 28531, 10178, 119, 17645, 6734, 27662, 14245, 35878, 9249, 1546, 46, 18547, 27408, 5087, 94, 51082], "temperature": 0, "avg_logprob": -0.09218816334390875, "compression_ratio": 1.239010989010989, "no_speech_prob": 1.1476121268561013e-11}, {"id": 172, "seek": 33056, "start": 344.9, "end": 347.12, "text": "才能打造出一个完整的解决方案", "tokens": [51082, 18888, 8225, 12467, 37583, 7781, 20182, 14128, 27662, 1546, 17278, 5676, 111, 9249, 28899, 51193], "temperature": 0, "avg_logprob": -0.09218816334390875, "compression_ratio": 1.239010989010989, "no_speech_prob": 1.1476121268561013e-11}, {"id": 173, "seek": 33056, "start": 347.12, "end": 348.16, "text": "总结来说", "tokens": [51193, 33440, 45641, 6912, 8090, 51245], "temperature": 0, "avg_logprob": -0.09218816334390875, "compression_ratio": 1.239010989010989, "no_speech_prob": 1.1476121268561013e-11}, {"id": 174, "seek": 33056, "start": 348.16, "end": 350.72, "text": "PDF Inspector并不是又一个单纯的", "tokens": [51245, 17349, 37, 33402, 3509, 114, 7296, 17047, 20182, 47446, 16853, 107, 1546, 51373], "temperature": 0, "avg_logprob": -0.09218816334390875, "compression_ratio": 1.239010989010989, "no_speech_prob": 1.1476121268561013e-11}, {"id": 175, "seek": 33056, "start": 350.72, "end": 351.9, "text": "PDF转档工具", "tokens": [51373, 17349, 37, 17819, 105, 21416, 96, 23323, 39806, 51432], "temperature": 0, "avg_logprob": -0.09218816334390875, "compression_ratio": 1.239010989010989, "no_speech_prob": 1.1476121268561013e-11}, {"id": 176, "seek": 33056, "start": 351.9, "end": 354.66, "text": "它更像是一个为了现代化文件处理流程", "tokens": [51432, 11284, 19002, 12760, 1541, 20182, 13992, 2289, 20204, 19105, 23756, 17174, 20485, 1787, 226, 13876, 27854, 29649, 51570], "temperature": 0, "avg_logprob": -0.09218816334390875, "compression_ratio": 1.239010989010989, "no_speech_prob": 1.1476121268561013e-11}, {"id": 177, "seek": 33056, "start": 354.66, "end": 356.96, "text": "所设计的智慧调度核心", "tokens": [51570, 5966, 7422, 122, 7422, 94, 1546, 5094, 118, 12358, 100, 8897, 225, 13127, 8987, 116, 7945, 51685], "temperature": 0, "avg_logprob": -0.09218816334390875, "compression_ratio": 1.239010989010989, "no_speech_prob": 1.1476121268561013e-11}, {"id": 178, "seek": 33056, "start": 356.96, "end": 357.98, "text": "它的价值", "tokens": [51685, 45224, 1550, 115, 40242, 51736], "temperature": 0, "avg_logprob": -0.09218816334390875, "compression_ratio": 1.239010989010989, "no_speech_prob": 1.1476121268561013e-11}, {"id": 179, "seek": 33056, "start": 357.98, "end": 359.96, "text": "并不是要取代OCR", "tokens": [51736, 3509, 114, 7296, 4275, 29436, 19105, 46, 18547, 51835], "temperature": 0, "avg_logprob": -0.09218816334390875, "compression_ratio": 1.239010989010989, "no_speech_prob": 1.1476121268561013e-11}, {"id": 180, "seek": 35996, "start": 359.96, "end": 361.71999999999997, "text": "而是要让OCR的使用", "tokens": [50365, 11070, 1541, 4275, 33650, 46, 18547, 1546, 22982, 9254, 50453], "temperature": 0, "avg_logprob": -0.08167017050806036, "compression_ratio": 1.2359249329758712, "no_speech_prob": 1.2355725817481122e-11}, {"id": 181, "seek": 35996, "start": 361.71999999999997, "end": 362.9, "text": "变得更聪明", "tokens": [50453, 2129, 246, 5916, 19002, 8171, 103, 11100, 50512], "temperature": 0, "avg_logprob": -0.08167017050806036, "compression_ratio": 1.2359249329758712, "no_speech_prob": 1.2355725817481122e-11}, {"id": 182, "seek": 35996, "start": 362.9, "end": 363.82, "text": "更有效率", "tokens": [50512, 19002, 2412, 43076, 44866, 50558], "temperature": 0, "avg_logprob": -0.08167017050806036, "compression_ratio": 1.2359249329758712, "no_speech_prob": 1.2355725817481122e-11}, {"id": 183, "seek": 35996, "start": 363.82, "end": 365.94, "text": "对于任何需要大规模处理", "tokens": [50558, 8713, 37732, 26443, 11906, 35748, 3582, 6758, 226, 41908, 1787, 226, 13876, 50664], "temperature": 0, "avg_logprob": -0.08167017050806036, "compression_ratio": 1.2359249329758712, "no_speech_prob": 1.2355725817481122e-11}, {"id": 184, "seek": 35996, "start": 365.94, "end": 367.88, "text": "PDF的开发者或组织来说", "tokens": [50664, 17349, 37, 1546, 18937, 28926, 12444, 19780, 10115, 226, 10115, 229, 6912, 8090, 50761], "temperature": 0, "avg_logprob": -0.08167017050806036, "compression_ratio": 1.2359249329758712, "no_speech_prob": 1.2355725817481122e-11}, {"id": 185, "seek": 35996, "start": 367.88, "end": 370.08, "text": "这个专案最值得关注的地方", "tokens": [50761, 15368, 940, 241, 28899, 8661, 40242, 5916, 28053, 26432, 1546, 30146, 50871], "temperature": 0, "avg_logprob": -0.08167017050806036, "compression_ratio": 1.2359249329758712, "no_speech_prob": 1.2355725817481122e-11}, {"id": 186, "seek": 35996, "start": 370.08, "end": 372.78, "text": "就是它提供了一个绝佳的机会", "tokens": [50871, 5620, 11284, 20949, 3254, 249, 2289, 20182, 10115, 251, 1593, 111, 1546, 37960, 12949, 51006], "temperature": 0, "avg_logprob": -0.08167017050806036, "compression_ratio": 1.2359249329758712, "no_speech_prob": 1.2355725817481122e-11}, {"id": 187, "seek": 35996, "start": 372.78, "end": 374.85999999999996, "text": "让你用极低的整合成本", "tokens": [51006, 33650, 2166, 9254, 7360, 223, 41377, 1546, 27662, 14245, 11336, 8802, 51110], "temperature": 0, "avg_logprob": -0.08167017050806036, "compression_ratio": 1.2359249329758712, "no_speech_prob": 1.2355725817481122e-11}, {"id": 188, "seek": 35996, "start": 374.85999999999996, "end": 376.4, "text": "去重新检视", "tokens": [51110, 6734, 12624, 12560, 16407, 222, 40656, 51187], "temperature": 0, "avg_logprob": -0.08167017050806036, "compression_ratio": 1.2359249329758712, "no_speech_prob": 1.2355725817481122e-11}, {"id": 189, "seek": 35996, "start": 376.4, "end": 378.76, "text": "并且优化你整个系统里", "tokens": [51187, 3509, 114, 20334, 7384, 246, 23756, 2166, 27662, 7549, 25368, 10115, 253, 15759, 51305], "temperature": 0, "avg_logprob": -0.08167017050806036, "compression_ratio": 1.2359249329758712, "no_speech_prob": 1.2355725817481122e-11}, {"id": 190, "seek": 35996, "start": 378.76, "end": 380.24, "text": "那个最花钱的环节", "tokens": [51305, 43083, 8661, 20127, 39623, 1546, 8051, 107, 45161, 51379], "temperature": 0, "avg_logprob": -0.08167017050806036, "compression_ratio": 1.2359249329758712, "no_speech_prob": 1.2355725817481122e-11}, {"id": 191, "seek": 35996, "start": 380.24, "end": 381.79999999999995, "text": "以上就是今天的GitCoverty", "tokens": [51379, 29497, 5620, 34947, 38, 270, 34, 3570, 874, 51457], "temperature": 0, "avg_logprob": -0.08167017050806036, "compression_ratio": 1.2359249329758712, "no_speech_prob": 1.2355725817481122e-11}, {"id": 192, "seek": 35996, "start": 381.79999999999995, "end": 383.35999999999996, "text": "希望这些GitHub Trending专案", "tokens": [51457, 29955, 5562, 13824, 38, 270, 21150, 37417, 278, 940, 241, 28899, 51535], "temperature": 0, "avg_logprob": -0.08167017050806036, "compression_ratio": 1.2359249329758712, "no_speech_prob": 1.2355725817481122e-11}, {"id": 193, "seek": 35996, "start": 383.35999999999996, "end": 385.26, "text": "能为你的开发工作带来灵感", "tokens": [51535, 8225, 13992, 18961, 18937, 28926, 41315, 4845, 99, 6912, 13290, 113, 9709, 51630], "temperature": 0, "avg_logprob": -0.08167017050806036, "compression_ratio": 1.2359249329758712, "no_speech_prob": 1.2355725817481122e-11}, {"id": 194, "seek": 35996, "start": 385.26, "end": 386.59999999999997, "text": "以上这些GitHub Report", "tokens": [51630, 29497, 5562, 13824, 38, 270, 21150, 16057, 51697], "temperature": 0, "avg_logprob": -0.08167017050806036, "compression_ratio": 1.2359249329758712, "no_speech_prob": 1.2355725817481122e-11}, {"id": 195, "seek": 35996, "start": 386.59999999999997, "end": 388.03999999999996, "text": "请见下方资讯栏", "tokens": [51697, 27908, 23813, 4438, 9249, 5266, 226, 7422, 107, 8987, 237, 51769], "temperature": 0, "avg_logprob": -0.08167017050806036, "compression_ratio": 1.2359249329758712, "no_speech_prob": 1.2355725817481122e-11}, {"id": 196, "seek": 35996, "start": 388.03999999999996, "end": 389.21999999999997, "text": "如果你喜欢这个节目", "tokens": [51769, 45669, 41606, 15368, 45161, 11386, 51828], "temperature": 0, "avg_logprob": -0.08167017050806036, "compression_ratio": 1.2359249329758712, "no_speech_prob": 1.2355725817481122e-11}, {"id": 197, "seek": 38922, "start": 389.22, "end": 390.84000000000003, "text": "别忘了订阅我们的频道", "tokens": [50365, 18453, 26677, 2289, 7422, 95, 10034, 227, 15003, 1546, 39752, 6025, 50446], "temperature": 0, "avg_logprob": -0.11424851417541504, "compression_ratio": 0.8695652173913043, "no_speech_prob": 1.4046671811818179e-11}, {"id": 198, "seek": 38922, "start": 390.84000000000003, "end": 392.8, "text": "我们明天同一时间再见", "tokens": [50446, 15003, 11100, 6135, 13089, 2257, 44848, 44176, 50544], "temperature": 0, "avg_logprob": -0.11424851417541504, "compression_ratio": 0.8695652173913043, "no_speech_prob": 1.4046671811818179e-11}, {"id": 199, "seek": 39280, "start": 392.8, "end": 393.09000000000003, "text": "", "tokens": [], "temperature": 0, "avg_logprob": -0.7207106351852417, "compression_ratio": 0.0, "no_speech_prob": 6.266080432304477e-10, "words": []}], "language": "zh"}