{ "text": "号称agent 性能暴涨的Deep-Sea V4正式版模型\n在全面官宣兼容了Codex之后\n双键合并到底能开发出什么样惊艳的agent呢\n大家现在看到的就是这套技术方案真实的开发效果\n也是本期科快课最终的开发成果\n这是一套完整的工业级AI数据分析系统\n不仅能够实现全链路图文并茂的可视化数据分析\n以及机器学习建模预测分析\n并且还能将分析过程和关键结论一键整合\n并创建图文并茂的数据分析报告\n并且支持多版本迭代修改和一键回溯\n同时还支持海量数据库和本地数据的读取和导入\n项目里面以300万数据规模为例\n搭建了完整的数据分析流程\n此外agent还内置了大量的数据分析Skills\n用于支持数据探索可视化分析机器学习建模\n数据洞察和数据分析报告编写等各个环节\n并且支持Skills定制化开发和载入\n此外还能够实施查询系统状态完整架构以及各关键环节的底层代码\n甚至还融合了定音间的WebDesign设计理念\n主题搭配样式都能够一键切换\n怎么样是不是非常的酷炫\n说实话作为几年来一直奋战在技术一线的技术人\n看到国产模型能够达到如此效果还是非常的吃惊\n同时也非常兴奋\n因此本期科学课就为大家从零上手学习\nDeepSeek V4正式版加Kojax工业级agent开发技术完整体系\n详细介绍DeepSeek V4正式版模型的核心功能特性\n以及工业agent开发技术战\n当然DeepSeek V4能够全面接入Kojax\n并借助Kojax最大程度释放模型性能\n最核心的一点还是在于\n全新模型全面适配了OpenAI最新的Responses API\n不同于上一代Chat Compilations API\n核心功能是灵活编排消息队列\nResponses API更像是OpenAI版的LongChain\n开发者仅需要把模型系统提示词和当前任务放在一起\nResponses API就能够自动搭建起agent loop\n自动多步或者并发调用工具来执行任务\n也就是说上一代Chat Compilations API\n更加适合通过编排消息队列来搭建ChatBot聊天机器人\n而新一代的Responses API则更加适合开发agent\n这也怪不得OpenAI在Responses API迁移指南中说到\n仅仅是从Chat Compilations API迁移到了Responses API\nagent在SWE Bench性能方面就提升了3%\n而Deepseek也是目前唯一一个国内全面接入Responses API的大模型\n要知道哪怕在前段时间开源的2.8T参数的Kimi K3\n也仅仅支持上一代Chat Compilations API\n因此公开课的第一部分\n我就将为大家深入介绍\nDeepseek全新兼容的Responses API的基本语法规则和使用方法\n包括底层执行和调用逻辑\n帮大家快速奠定扎实基础\n而在了解了底层API之后\n接下来就进入到Kodex的上手取操环节\n公开课中将详细介绍Deepseek V4接入Kodex的具体方法\n包括脚本接入和CC Switch接入这两种方式\n同时详细介绍新人入门Kodex进行开发必备的基础技能\n包括基础环境搭建\n三层技艺管理系统方法\n多agent和profile设计\n工具链和写杠命令\nSkills与MCP工具调用等5个最核心的功能模块\n当然为了真实介入工业级开发流程\n我们还需要给Deepseek加Kodex这套技术体系\n接上视觉能力\n才能够在真实开发过程中实现开发成果的\n视觉自我审核外加审美配合对齐\n简单来说\n就是让agent通过看到自己的开发成果\n来提升开发效果\n国外科中将采用1013.8作为视觉工具\n并借助MCP来提供触发接口\n完整架构如图所示\n而Deepseek V4也只有在接入视觉能力之后\n才能够实现如图所示的非常惊艳的开发效果\n而在完成了一系列的基础准备工作之后\n工坏课的下半场\n就让我们进入工业级agent的开发实战阶段\n本期工坏课\n我们将同样选取目前最热门的agent的应用方向\nAI数据分析领域\n详细介绍\n如何搭建一整套工业级可用的AI数据分析系统\n其实经历了这么多年的发展\n工业场景下\n对于数据分析智能体\n早就不再是简单的提个数\n绘的图进行数据洞察分析\n得出几条结论这么简单的需求了\n而是首先要能接入各种数据员\n并至少支持百万级以上数据集的管理跟读取\n同时还要支持数据信息\n可视化分析\n数据洞察\n特征工程\n模型融合\n机器学习建模分析等完整链路\n以及还要能一键创建\n满足具体业务格式和指标要求的数据分析报告\n此外还要有SKILLS功能系统\n支持上传导入跟使用自定义的SKILL\n从而更好的进行功能拓展\n而这也是本期公开课中\n即将为大家深入系统讲解的\nDS4 AI数据分析系统\n作为教学项目\n我会从头\n为大家详细介绍AI数据分析的\n基本业务背景\n核心项目架构\n以及完整的技术选进方案\n并借助Codex一步步完成项目开发\n此外\n为了便于同学们之后\n自己也能熟练操作Codex\n完成自己的项目开发\n公开课中\n还特地为大家准备了一份\n完整的DeepSeek V4\n技术体系LMWiki知识库\n以及10项DeepSeek V4开发专属SKILLS\n目前全能公开课的课件\n脚本项目提示词\n知识库和SKILLS的\n都已经上限至复返AI社区\n加扫标评商的奥运码\n即可免费领取\n2016大模型Agent的技术发展\n是都非常迅猛\n我也将持续为大家提供\n最前沿实用的技术教学\n感谢大家的关注和三连支持\n公开课\n马上开始", "language": null, "duration": 326.33325, "segments": [ { "id": 0, "seek": 0, "start": 0.0, "end": 3.0, "text": "号称agent 性能暴涨的Deep-Sea V4正式版模型", "tokens": [ 50365, 26987, 8204, 108, 559, 317, 220, 21686, 8225, 23826, 112, 35681, 101, 1546, 11089, 595, 12, 10637, 64, 691, 19, 15789, 27584, 42096, 41908, 39823, 50515 ], "temperature": 0, "avg_logprob": -0.1449799601236979, "compression_ratio": 1.2921052631578946, "no_speech_prob": 3.033540085795039e-11, "source": "whisper" }, { "id": 1, "seek": 0, "start": 3.0, "end": 5.44, "text": "在全面官宣兼容了Codex之后", "tokens": [ 50515, 3581, 11319, 8833, 31929, 2415, 96, 2347, 120, 25750, 2289, 34, 1429, 87, 9574, 13547, 50637 ], "temperature": 0, "avg_logprob": -0.1449799601236979, "compression_ratio": 1.2921052631578946, "no_speech_prob": 3.033540085795039e-11, "source": "whisper" }, { "id": 2, "seek": 0, "start": 5.44, "end": 9.120000000000001, "text": "双键合并到底能开发出什么样惊艳的agent呢", "tokens": [ 50637, 2129, 234, 23049, 106, 14245, 3509, 114, 33883, 8225, 18937, 28926, 7781, 10440, 14496, 4799, 232, 12531, 111, 1546, 559, 317, 6240, 50821 ], "temperature": 0, "avg_logprob": -0.1449799601236979, "compression_ratio": 1.2921052631578946, "no_speech_prob": 3.033540085795039e-11, "source": "whisper" }, { "id": 3, "seek": 0, "start": 9.120000000000001, "end": 12.68, "text": "大家现在看到的就是这套技术方案真实的开发效果", "tokens": [ 50821, 6868, 25040, 18032, 1546, 5620, 5562, 1881, 245, 32502, 1474, 107, 9249, 28899, 6303, 24726, 1546, 18937, 28926, 43076, 9319, 50999 ], "temperature": 0, "avg_logprob": -0.1449799601236979, "compression_ratio": 1.2921052631578946, "no_speech_prob": 3.033540085795039e-11, "source": "whisper" }, { "id": 4, "seek": 0, "start": 12.68, "end": 15.0, "text": "也是本期科快课最终的开发成果", "tokens": [ 50999, 22021, 8802, 16786, 42091, 10251, 5233, 122, 8661, 10115, 230, 1546, 18937, 28926, 11336, 9319, 51115 ], "temperature": 0, "avg_logprob": -0.1449799601236979, "compression_ratio": 1.2921052631578946, "no_speech_prob": 3.033540085795039e-11, "source": "whisper" }, { "id": 5, "seek": 0, "start": 15.0, "end": 17.72, "text": "这是一套完整的工业级AI数据分析系统", "tokens": [ 51115, 27455, 2257, 1881, 245, 14128, 27662, 1546, 23323, 940, 248, 16853, 100, 48698, 33188, 26075, 106, 6627, 7360, 238, 25368, 10115, 253, 51251 ], "temperature": 0, "avg_logprob": -0.1449799601236979, "compression_ratio": 1.2921052631578946, "no_speech_prob": 3.033540085795039e-11, "source": "whisper" }, { "id": 6, "seek": 0, "start": 17.72, "end": 21.240000000000002, "text": "不仅能够实现全链路图文并茂的可视化数据分析", "tokens": [ 51251, 1960, 1550, 227, 8225, 1787, 253, 24726, 20204, 11319, 165, 241, 122, 24658, 3919, 122, 17174, 3509, 114, 27395, 224, 1546, 4429, 40656, 23756, 33188, 26075, 106, 6627, 7360, 238, 51427 ], "temperature": 0, "avg_logprob": -0.1449799601236979, "compression_ratio": 1.2921052631578946, "no_speech_prob": 3.033540085795039e-11, "source": "whisper" }, { "id": 7, "seek": 0, "start": 21.240000000000002, "end": 23.2, "text": "以及机器学习建模预测分析", "tokens": [ 51427, 40282, 37960, 34386, 29618, 2930, 254, 34157, 41908, 12501, 226, 11038, 233, 6627, 7360, 238, 51525 ], "temperature": 0, "avg_logprob": -0.1449799601236979, "compression_ratio": 1.2921052631578946, "no_speech_prob": 3.033540085795039e-11, "source": "whisper" }, { "id": 8, "seek": 0, "start": 23.2, "end": 26.52, "text": "并且还能将分析过程和关键结论一键整合", "tokens": [ 51525, 3509, 114, 20334, 14852, 8225, 45456, 6627, 7360, 238, 16866, 29649, 12565, 28053, 23049, 106, 45641, 7422, 118, 2257, 23049, 106, 27662, 14245, 51691 ], "temperature": 0, "avg_logprob": -0.1449799601236979, "compression_ratio": 1.2921052631578946, "no_speech_prob": 3.033540085795039e-11, "source": "whisper" }, { "id": 9, "seek": 2652, "start": 26.52, "end": 28.68, "text": "并创建图文并茂的数据分析报告", "tokens": [ 50365, 3509, 114, 2437, 249, 34157, 3919, 122, 17174, 3509, 114, 27395, 224, 1546, 33188, 26075, 106, 6627, 7360, 238, 49817, 16846, 50473 ], "temperature": 0, "avg_logprob": -0.06276125166151258, "compression_ratio": 1.394736842105263, "no_speech_prob": 1.2715990586886772e-11, "source": "whisper" }, { "id": 10, "seek": 2652, "start": 28.68, "end": 32.12, "text": "并且支持多版本迭代修改和一键回溯", "tokens": [ 50473, 3509, 114, 20334, 35488, 6392, 42096, 8802, 3316, 255, 19105, 220, 7792, 106, 34490, 12565, 2257, 23049, 106, 8350, 16084, 107, 50645 ], "temperature": 0, "avg_logprob": -0.06276125166151258, "compression_ratio": 1.394736842105263, "no_speech_prob": 1.2715990586886772e-11, "source": "whisper" }, { "id": 11, "seek": 2652, "start": 32.12, "end": 35.8, "text": "同时还支持海量数据库和本地数据的读取和导入", "tokens": [ 50645, 13089, 15729, 14852, 35488, 20078, 26748, 33188, 26075, 106, 6346, 241, 12565, 8802, 10928, 33188, 26075, 106, 1546, 5233, 119, 29436, 12565, 4510, 120, 14028, 50829 ], "temperature": 0, "avg_logprob": -0.06276125166151258, "compression_ratio": 1.394736842105263, "no_speech_prob": 1.2715990586886772e-11, "source": "whisper" }, { "id": 12, "seek": 2652, "start": 35.8, "end": 37.980000000000004, "text": "项目里面以300万数据规模为例", "tokens": [ 50829, 10178, 117, 11386, 15759, 8833, 3588, 12566, 23570, 33188, 26075, 106, 6758, 226, 41908, 13992, 17797, 50938 ], "temperature": 0, "avg_logprob": -0.06276125166151258, "compression_ratio": 1.394736842105263, "no_speech_prob": 1.2715990586886772e-11, "source": "whisper" }, { "id": 13, "seek": 2652, "start": 37.980000000000004, "end": 39.58, "text": "搭建了完整的数据分析流程", "tokens": [ 50938, 22220, 255, 34157, 2289, 14128, 27662, 1546, 33188, 26075, 106, 6627, 7360, 238, 27854, 29649, 51018 ], "temperature": 0, "avg_logprob": -0.06276125166151258, "compression_ratio": 1.394736842105263, "no_speech_prob": 1.2715990586886772e-11, "source": "whisper" }, { "id": 14, "seek": 2652, "start": 39.58, "end": 42.58, "text": "此外agent还内置了大量的数据分析Skills", "tokens": [ 51018, 17947, 12022, 559, 317, 14852, 34742, 34719, 2289, 3582, 26748, 1546, 33188, 26075, 106, 6627, 7360, 238, 50, 74, 2565, 51168 ], "temperature": 0, "avg_logprob": -0.06276125166151258, "compression_ratio": 1.394736842105263, "no_speech_prob": 1.2715990586886772e-11, "source": "whisper" }, { "id": 15, "seek": 2652, "start": 42.58, "end": 45.84, "text": "用于支持数据探索可视化分析机器学习建模", "tokens": [ 51168, 9254, 37732, 35488, 33188, 26075, 106, 6900, 95, 7732, 95, 220, 4429, 40656, 23756, 6627, 7360, 238, 220, 37960, 34386, 29618, 2930, 254, 34157, 41908, 51331 ], "temperature": 0, "avg_logprob": -0.06276125166151258, "compression_ratio": 1.394736842105263, "no_speech_prob": 1.2715990586886772e-11, "source": "whisper" }, { "id": 16, "seek": 2652, "start": 45.84, "end": 48.92, "text": "数据洞察和数据分析报告编写等各个环节", "tokens": [ 51331, 33188, 26075, 106, 12136, 252, 47550, 12565, 33188, 26075, 106, 6627, 7360, 238, 49817, 16846, 38109, 244, 5676, 247, 10187, 17516, 7549, 8051, 107, 45161, 51485 ], "temperature": 0, "avg_logprob": -0.06276125166151258, "compression_ratio": 1.394736842105263, "no_speech_prob": 1.2715990586886772e-11, "source": "whisper" }, { "id": 17, "seek": 2652, "start": 48.92, "end": 51.96, "text": "并且支持Skills定制化开发和载入", "tokens": [ 51485, 3509, 114, 20334, 35488, 50, 74, 2565, 12088, 25491, 23756, 18937, 28926, 12565, 17819, 121, 14028, 51637 ], "temperature": 0, "avg_logprob": -0.06276125166151258, "compression_ratio": 1.394736842105263, "no_speech_prob": 1.2715990586886772e-11, "source": "whisper" }, { "id": 18, "seek": 5196, "start": 51.96, "end": 57.06, "text": "此外还能够实施查询系统状态完整架构以及各关键环节的底层代码", "tokens": [ 50365, 17947, 12022, 14852, 8225, 1787, 253, 24726, 4307, 121, 42623, 5233, 95, 25368, 10115, 253, 35276, 114, 3757, 223, 220, 14128, 27662, 7360, 114, 7360, 226, 40282, 17516, 28053, 23049, 106, 8051, 107, 45161, 1546, 22816, 9636, 224, 19105, 23230, 223, 50620 ], "temperature": 0, "avg_logprob": -0.11979291085421183, "compression_ratio": 1.2299741602067182, "no_speech_prob": 1.6495704108221432e-11, "source": "whisper" }, { "id": 19, "seek": 5196, "start": 57.06, "end": 60.7, "text": "甚至还融合了定音间的WebDesign设计理念", "tokens": [ 50620, 17556, 20844, 14852, 43772, 235, 14245, 2289, 12088, 18034, 31685, 1546, 4360, 65, 29543, 788, 7422, 122, 7422, 94, 13876, 33679, 50802 ], "temperature": 0, "avg_logprob": -0.11979291085421183, "compression_ratio": 1.2299741602067182, "no_speech_prob": 1.6495704108221432e-11, "source": "whisper" }, { "id": 20, "seek": 5196, "start": 60.7, "end": 63.42, "text": "主题搭配样式都能够一键切换", "tokens": [ 50802, 13557, 30716, 22220, 255, 38846, 14496, 27584, 7182, 8225, 1787, 253, 2257, 23049, 106, 23632, 26075, 95, 50938 ], "temperature": 0, "avg_logprob": -0.11979291085421183, "compression_ratio": 1.2299741602067182, "no_speech_prob": 1.6495704108221432e-11, "source": "whisper" }, { "id": 21, "seek": 5196, "start": 63.42, "end": 65.16, "text": "怎么样是不是非常的酷炫", "tokens": [ 50938, 48200, 23034, 48263, 16911, 115, 4804, 104, 51025 ], "temperature": 0, "avg_logprob": -0.11979291085421183, "compression_ratio": 1.2299741602067182, "no_speech_prob": 1.6495704108221432e-11, "source": "whisper" }, { "id": 22, "seek": 5196, "start": 65.16, "end": 69.2, "text": "说实话作为几年来一直奋战在技术一线的技术人", "tokens": [ 51025, 8090, 24726, 21596, 220, 11914, 13992, 6336, 254, 5157, 6912, 34448, 1881, 233, 1486, 246, 3581, 32502, 1474, 107, 2257, 16853, 123, 1546, 32502, 1474, 107, 4035, 51227 ], "temperature": 0, "avg_logprob": -0.11979291085421183, "compression_ratio": 1.2299741602067182, "no_speech_prob": 1.6495704108221432e-11, "source": "whisper" }, { "id": 23, "seek": 5196, "start": 69.2, "end": 72.74000000000001, "text": "看到国产模型能够达到如此效果还是非常的吃惊", "tokens": [ 51227, 18032, 16086, 1369, 100, 41908, 39823, 8225, 1787, 253, 9830, 122, 4511, 48487, 43076, 9319, 45726, 48263, 10123, 4799, 232, 51404 ], "temperature": 0, "avg_logprob": -0.11979291085421183, "compression_ratio": 1.2299741602067182, "no_speech_prob": 1.6495704108221432e-11, "source": "whisper" }, { "id": 24, "seek": 5196, "start": 72.74000000000001, "end": 73.94, "text": "同时也非常兴奋", "tokens": [ 51404, 13089, 15729, 6404, 14392, 2347, 112, 1881, 233, 51464 ], "temperature": 0, "avg_logprob": -0.11979291085421183, "compression_ratio": 1.2299741602067182, "no_speech_prob": 1.6495704108221432e-11, "source": "whisper" }, { "id": 25, "seek": 5196, "start": 73.94, "end": 76.92, "text": "因此本期科学课就为大家从零上手学习", "tokens": [ 51464, 8698, 17947, 8802, 16786, 42091, 29618, 5233, 122, 3111, 13992, 6868, 35630, 6306, 114, 5708, 11389, 29618, 2930, 254, 51613 ], "temperature": 0, "avg_logprob": -0.11979291085421183, "compression_ratio": 1.2299741602067182, "no_speech_prob": 1.6495704108221432e-11, "source": "whisper" }, { "id": 26, "seek": 5196, "start": 76.92, "end": 81.56, "text": "DeepSeek V4正式版加Kojax工业级agent开发技术完整体系", "tokens": [ 51613, 11089, 595, 10637, 916, 691, 19, 15789, 27584, 42096, 9990, 42, 37224, 2797, 23323, 940, 248, 16853, 100, 559, 317, 18937, 28926, 32502, 1474, 107, 14128, 27662, 29485, 25368, 51845 ], "temperature": 0, "avg_logprob": -0.11979291085421183, "compression_ratio": 1.2299741602067182, "no_speech_prob": 1.6495704108221432e-11, "source": "whisper" }, { "id": 27, "seek": 8156, "start": 81.56, "end": 84.9, "text": "详细介绍DeepSeek V4正式版模型的核心功能特性", "tokens": [ 50365, 5233, 99, 10115, 228, 30312, 10115, 235, 11089, 595, 10637, 916, 691, 19, 15789, 27584, 42096, 41908, 39823, 1546, 8987, 116, 7945, 32311, 8225, 17682, 21686, 50532 ], "temperature": 0, "avg_logprob": -0.09045394425539627, "compression_ratio": 1.274787535410765, "no_speech_prob": 1.498224287688199e-11, "source": "whisper" }, { "id": 28, "seek": 8156, "start": 84.9, "end": 86.94, "text": "以及工业agent开发技术战", "tokens": [ 50532, 40282, 23323, 940, 248, 559, 317, 18937, 28926, 32502, 1474, 107, 1486, 246, 50634 ], "temperature": 0, "avg_logprob": -0.09045394425539627, "compression_ratio": 1.274787535410765, "no_speech_prob": 1.498224287688199e-11, "source": "whisper" }, { "id": 29, "seek": 8156, "start": 86.94, "end": 90.02000000000001, "text": "当然DeepSeek V4能够全面接入Kojax", "tokens": [ 50634, 40486, 11089, 595, 10637, 916, 691, 19, 8225, 1787, 253, 11319, 8833, 14468, 14028, 42, 37224, 2797, 50788 ], "temperature": 0, "avg_logprob": -0.09045394425539627, "compression_ratio": 1.274787535410765, "no_speech_prob": 1.498224287688199e-11, "source": "whisper" }, { "id": 30, "seek": 8156, "start": 90.02000000000001, "end": 92.88, "text": "并借助Kojax最大程度释放模型性能", "tokens": [ 50788, 3509, 114, 2122, 253, 37618, 42, 37224, 2797, 8661, 3582, 29649, 13127, 5873, 232, 12744, 41908, 39823, 21686, 8225, 50931 ], "temperature": 0, "avg_logprob": -0.09045394425539627, "compression_ratio": 1.274787535410765, "no_speech_prob": 1.498224287688199e-11, "source": "whisper" }, { "id": 31, "seek": 8156, "start": 92.88, "end": 94.86, "text": "最核心的一点还是在于", "tokens": [ 50931, 8661, 8987, 116, 7945, 1546, 49042, 45726, 3581, 37732, 51030 ], "temperature": 0, "avg_logprob": -0.09045394425539627, "compression_ratio": 1.274787535410765, "no_speech_prob": 1.498224287688199e-11, "source": "whisper" }, { "id": 32, "seek": 8156, "start": 94.86, "end": 98.44, "text": "全新模型全面适配了OpenAI最新的Responses API", "tokens": [ 51030, 11319, 12560, 41908, 39823, 11319, 8833, 2215, 224, 38846, 2289, 45569, 48698, 8661, 12560, 1546, 49, 13361, 892, 279, 9362, 51209 ], "temperature": 0, "avg_logprob": -0.09045394425539627, "compression_ratio": 1.274787535410765, "no_speech_prob": 1.498224287688199e-11, "source": "whisper" }, { "id": 33, "seek": 8156, "start": 98.44, "end": 100.98, "text": "不同于上一代Chat Compilations API", "tokens": [ 51209, 47123, 37732, 5708, 2257, 19105, 41683, 6620, 388, 763, 9362, 51336 ], "temperature": 0, "avg_logprob": -0.09045394425539627, "compression_ratio": 1.274787535410765, "no_speech_prob": 1.498224287688199e-11, "source": "whisper" }, { "id": 34, "seek": 8156, "start": 100.98, "end": 103.0, "text": "核心功能是灵活编排消息队列", "tokens": [ 51336, 8987, 116, 7945, 32311, 8225, 1541, 13290, 113, 25956, 38109, 244, 44647, 28837, 26460, 10034, 253, 43338, 51437 ], "temperature": 0, "avg_logprob": -0.09045394425539627, "compression_ratio": 1.274787535410765, "no_speech_prob": 1.498224287688199e-11, "source": "whisper" }, { "id": 35, "seek": 8156, "start": 103.0, "end": 106.48, "text": "Responses API更像是OpenAI版的LongChain", "tokens": [ 51437, 49, 13361, 892, 279, 9362, 19002, 12760, 1541, 45569, 48698, 42096, 1546, 43, 556, 6546, 491, 51611 ], "temperature": 0, "avg_logprob": -0.09045394425539627, "compression_ratio": 1.274787535410765, "no_speech_prob": 1.498224287688199e-11, "source": "whisper" }, { "id": 36, "seek": 8156, "start": 106.48, "end": 110.74000000000001, "text": "开发者仅需要把模型系统提示词和当前任务放在一起", "tokens": [ 51611, 18937, 28926, 12444, 1550, 227, 35748, 16075, 41908, 39823, 25368, 10115, 253, 20949, 25696, 5233, 235, 12565, 16233, 8945, 26443, 5087, 94, 12744, 3581, 29567, 51824 ], "temperature": 0, "avg_logprob": -0.09045394425539627, "compression_ratio": 1.274787535410765, "no_speech_prob": 1.498224287688199e-11, "source": "whisper" }, { "id": 37, "seek": 11074, "start": 110.74, "end": 113.94, "text": "Responses API就能够自动搭建起agent loop", "tokens": [ 50365, 49, 13361, 892, 279, 9362, 3111, 8225, 1787, 253, 9722, 34961, 22220, 255, 34157, 9147, 559, 317, 6367, 50525 ], "temperature": 0, "avg_logprob": -0.08065681579785469, "compression_ratio": 1.3181818181818181, "no_speech_prob": 1.4219631548628708e-11, "source": "whisper" }, { "id": 38, "seek": 11074, "start": 113.94, "end": 117.39999999999999, "text": "自动多步或者并发调用工具来执行任务", "tokens": [ 50525, 9722, 34961, 6392, 31429, 31148, 3509, 114, 28926, 8897, 225, 9254, 23323, 39806, 6912, 3416, 100, 8082, 26443, 5087, 94, 50698 ], "temperature": 0, "avg_logprob": -0.08065681579785469, "compression_ratio": 1.3181818181818181, "no_speech_prob": 1.4219631548628708e-11, "source": "whisper" }, { "id": 39, "seek": 11074, "start": 117.39999999999999, "end": 120.0, "text": "也就是说上一代Chat Compilations API", "tokens": [ 50698, 6404, 5620, 8090, 5708, 2257, 19105, 41683, 6620, 388, 763, 9362, 50828 ], "temperature": 0, "avg_logprob": -0.08065681579785469, "compression_ratio": 1.3181818181818181, "no_speech_prob": 1.4219631548628708e-11, "source": "whisper" }, { "id": 40, "seek": 11074, "start": 120.0, "end": 124.17999999999999, "text": "更加适合通过编排消息队列来搭建ChatBot聊天机器人", "tokens": [ 50828, 19002, 9990, 2215, 224, 14245, 19550, 16866, 38109, 244, 44647, 28837, 26460, 10034, 253, 43338, 6912, 22220, 255, 34157, 41683, 33, 310, 40096, 6135, 37960, 34386, 4035, 51037 ], "temperature": 0, "avg_logprob": -0.08065681579785469, "compression_ratio": 1.3181818181818181, "no_speech_prob": 1.4219631548628708e-11, "source": "whisper" }, { "id": 41, "seek": 11074, "start": 124.17999999999999, "end": 127.89999999999999, "text": "而新一代的Responses API则更加适合开发agent", "tokens": [ 51037, 11070, 12560, 2257, 19105, 1546, 49, 13361, 892, 279, 9362, 2437, 247, 19002, 9990, 2215, 224, 14245, 18937, 28926, 559, 317, 51223 ], "temperature": 0, "avg_logprob": -0.08065681579785469, "compression_ratio": 1.3181818181818181, "no_speech_prob": 1.4219631548628708e-11, "source": "whisper" }, { "id": 42, "seek": 11074, "start": 127.89999999999999, "end": 131.98, "text": "这也怪不得OpenAI在Responses API迁移指南中说到", "tokens": [ 51223, 5562, 6404, 26798, 1960, 5916, 45569, 48698, 3581, 49, 13361, 892, 279, 9362, 3316, 223, 8204, 119, 25922, 31021, 5975, 8090, 4511, 51427 ], "temperature": 0, "avg_logprob": -0.08065681579785469, "compression_ratio": 1.3181818181818181, "no_speech_prob": 1.4219631548628708e-11, "source": "whisper" }, { "id": 43, "seek": 11074, "start": 131.98, "end": 136.4, "text": "仅仅是从Chat Compilations API迁移到了Responses API", "tokens": [ 51427, 1550, 227, 1550, 227, 1541, 35630, 41683, 6620, 388, 763, 9362, 3316, 223, 8204, 119, 21381, 49, 13361, 892, 279, 9362, 51648 ], "temperature": 0, "avg_logprob": -0.08065681579785469, "compression_ratio": 1.3181818181818181, "no_speech_prob": 1.4219631548628708e-11, "source": "whisper" }, { "id": 44, "seek": 13640, "start": 136.4, "end": 140.72, "text": "agent在SWE Bench性能方面就提升了3%", "tokens": [ 50365, 559, 317, 3581, 50, 37937, 3964, 339, 21686, 8225, 39604, 3111, 20949, 41670, 2289, 18, 4, 50581 ], "temperature": 0, "avg_logprob": -0.08969540266120958, "compression_ratio": 1.164179104477612, "no_speech_prob": 2.0303947759203744e-11, "source": "whisper" }, { "id": 45, "seek": 13640, "start": 140.72, "end": 146.02, "text": "而Deepseek也是目前唯一一个国内全面接入Responses API的大模型", "tokens": [ 50581, 11070, 11089, 595, 405, 916, 22021, 39004, 17198, 107, 2257, 20182, 16086, 34742, 11319, 8833, 14468, 14028, 49, 13361, 892, 279, 9362, 1546, 3582, 41908, 39823, 50846 ], "temperature": 0, "avg_logprob": -0.08969540266120958, "compression_ratio": 1.164179104477612, "no_speech_prob": 2.0303947759203744e-11, "source": "whisper" }, { "id": 46, "seek": 13640, "start": 146.02, "end": 150.6, "text": "要知道哪怕在前段时间开源的2.8T参数的Kimi K3", "tokens": [ 50846, 4275, 7758, 17028, 21164, 3581, 8945, 28427, 44848, 18937, 47402, 1546, 17, 13, 23, 51, 2129, 224, 33188, 1546, 42, 10121, 591, 18, 51075 ], "temperature": 0, "avg_logprob": -0.08969540266120958, "compression_ratio": 1.164179104477612, "no_speech_prob": 2.0303947759203744e-11, "source": "whisper" }, { "id": 47, "seek": 13640, "start": 150.6, "end": 153.6, "text": "也仅仅支持上一代Chat Compilations API", "tokens": [ 51075, 6404, 1550, 227, 1550, 227, 35488, 5708, 2257, 19105, 41683, 6620, 388, 763, 9362, 51225 ], "temperature": 0, "avg_logprob": -0.08969540266120958, "compression_ratio": 1.164179104477612, "no_speech_prob": 2.0303947759203744e-11, "source": "whisper" }, { "id": 48, "seek": 13640, "start": 153.6, "end": 154.98000000000002, "text": "因此公开课的第一部分", "tokens": [ 51225, 8698, 17947, 13545, 18937, 5233, 122, 1546, 18049, 32174, 51294 ], "temperature": 0, "avg_logprob": -0.08969540266120958, "compression_ratio": 1.164179104477612, "no_speech_prob": 2.0303947759203744e-11, "source": "whisper" }, { "id": 49, "seek": 13640, "start": 154.98000000000002, "end": 156.4, "text": "我就将为大家深入介绍", "tokens": [ 51294, 22020, 45456, 13992, 6868, 24043, 14028, 30312, 10115, 235, 51365 ], "temperature": 0, "avg_logprob": -0.08969540266120958, "compression_ratio": 1.164179104477612, "no_speech_prob": 2.0303947759203744e-11, "source": "whisper" }, { "id": 50, "seek": 13640, "start": 156.4, "end": 160.76, "text": "Deepseek全新兼容的Responses API的基本语法规则和使用方法", "tokens": [ 51365, 11089, 595, 405, 916, 11319, 12560, 2347, 120, 25750, 1546, 49, 13361, 892, 279, 9362, 1546, 37946, 5233, 255, 11148, 6758, 226, 2437, 247, 12565, 22982, 9254, 9249, 11148, 51583 ], "temperature": 0, "avg_logprob": -0.08969540266120958, "compression_ratio": 1.164179104477612, "no_speech_prob": 2.0303947759203744e-11, "source": "whisper" }, { "id": 51, "seek": 13640, "start": 160.76, "end": 163.16, "text": "包括底层执行和调用逻辑", "tokens": [ 51583, 41828, 22816, 9636, 224, 3416, 100, 8082, 12565, 8897, 225, 9254, 2215, 119, 9830, 239, 51703 ], "temperature": 0, "avg_logprob": -0.08969540266120958, "compression_ratio": 1.164179104477612, "no_speech_prob": 2.0303947759203744e-11, "source": "whisper" }, { "id": 52, "seek": 16316, "start": 163.16, "end": 165.6, "text": "帮大家快速奠定扎实基础", "tokens": [ 50365, 4845, 106, 6868, 10251, 31217, 1881, 254, 12088, 3416, 236, 24726, 26008, 38114, 222, 50487 ], "temperature": 0, "avg_logprob": -0.1077792207959672, "compression_ratio": 1.2096774193548387, "no_speech_prob": 1.6287589332808494e-11, "source": "whisper" }, { "id": 53, "seek": 16316, "start": 165.6, "end": 167.34, "text": "而在了解了底层API之后", "tokens": [ 50487, 11070, 3581, 2289, 17278, 2289, 22816, 9636, 224, 4715, 40, 9574, 13547, 50574 ], "temperature": 0, "avg_logprob": -0.1077792207959672, "compression_ratio": 1.2096774193548387, "no_speech_prob": 1.6287589332808494e-11, "source": "whisper" }, { "id": 54, "seek": 16316, "start": 167.34, "end": 169.56, "text": "接下来就进入到Kodex的上手取操环节", "tokens": [ 50574, 14468, 4438, 6912, 3111, 36700, 14028, 4511, 42, 1429, 87, 1546, 5708, 11389, 29436, 13167, 235, 8051, 107, 45161, 50685 ], "temperature": 0, "avg_logprob": -0.1077792207959672, "compression_ratio": 1.2096774193548387, "no_speech_prob": 1.6287589332808494e-11, "source": "whisper" }, { "id": 55, "seek": 16316, "start": 169.56, "end": 173.74, "text": "公开课中将详细介绍Deepseek V4接入Kodex的具体方法", "tokens": [ 50685, 13545, 18937, 5233, 122, 5975, 45456, 5233, 99, 10115, 228, 30312, 10115, 235, 11089, 595, 405, 916, 691, 19, 14468, 14028, 42, 1429, 87, 1546, 39806, 29485, 9249, 11148, 50894 ], "temperature": 0, "avg_logprob": -0.1077792207959672, "compression_ratio": 1.2096774193548387, "no_speech_prob": 1.6287589332808494e-11, "source": "whisper" }, { "id": 56, "seek": 16316, "start": 173.74, "end": 177.24, "text": "包括脚本接入和CC Switch接入这两种方式", "tokens": [ 50894, 41828, 27067, 248, 8802, 14468, 14028, 12565, 11717, 13893, 14468, 14028, 5562, 36257, 39810, 9249, 27584, 51069 ], "temperature": 0, "avg_logprob": -0.1077792207959672, "compression_ratio": 1.2096774193548387, "no_speech_prob": 1.6287589332808494e-11, "source": "whisper" }, { "id": 57, "seek": 16316, "start": 177.24, "end": 181.94, "text": "同时详细介绍新人入门Kodex进行开发必备的基础技能", "tokens": [ 51069, 13089, 15729, 5233, 99, 10115, 228, 30312, 10115, 235, 12560, 4035, 14028, 8259, 101, 42, 1429, 87, 36700, 8082, 18937, 28926, 28531, 1787, 229, 1546, 26008, 38114, 222, 32502, 8225, 51304 ], "temperature": 0, "avg_logprob": -0.1077792207959672, "compression_ratio": 1.2096774193548387, "no_speech_prob": 1.6287589332808494e-11, "source": "whisper" }, { "id": 58, "seek": 16316, "start": 181.94, "end": 183.46, "text": "包括基础环境搭建", "tokens": [ 51304, 41828, 26008, 38114, 222, 8051, 107, 47885, 22220, 255, 34157, 51380 ], "temperature": 0, "avg_logprob": -0.1077792207959672, "compression_ratio": 1.2096774193548387, "no_speech_prob": 1.6287589332808494e-11, "source": "whisper" }, { "id": 59, "seek": 16316, "start": 183.46, "end": 185.3, "text": "三层技艺管理系统方法", "tokens": [ 51380, 10960, 9636, 224, 32502, 12531, 118, 23131, 13876, 25368, 10115, 253, 9249, 11148, 51472 ], "temperature": 0, "avg_logprob": -0.1077792207959672, "compression_ratio": 1.2096774193548387, "no_speech_prob": 1.6287589332808494e-11, "source": "whisper" }, { "id": 60, "seek": 16316, "start": 185.3, "end": 187.35999999999999, "text": "多agent和profile设计", "tokens": [ 51472, 6392, 559, 317, 12565, 29175, 794, 7422, 122, 7422, 94, 51575 ], "temperature": 0, "avg_logprob": -0.1077792207959672, "compression_ratio": 1.2096774193548387, "no_speech_prob": 1.6287589332808494e-11, "source": "whisper" }, { "id": 61, "seek": 16316, "start": 187.35999999999999, "end": 188.76, "text": "工具链和写杠命令", "tokens": [ 51575, 23323, 39806, 165, 241, 122, 12565, 5676, 247, 4422, 254, 25236, 49061, 51645 ], "temperature": 0, "avg_logprob": -0.1077792207959672, "compression_ratio": 1.2096774193548387, "no_speech_prob": 1.6287589332808494e-11, "source": "whisper" }, { "id": 62, "seek": 16316, "start": 188.76, "end": 192.66, "text": "Skills与MCP工具调用等5个最核心的功能模块", "tokens": [ 51645, 50, 74, 2565, 940, 236, 44, 20049, 23323, 39806, 8897, 225, 9254, 10187, 20, 7549, 8661, 8987, 116, 7945, 1546, 32311, 8225, 41908, 47734, 51840 ], "temperature": 0, "avg_logprob": -0.1077792207959672, "compression_ratio": 1.2096774193548387, "no_speech_prob": 1.6287589332808494e-11, "source": "whisper" }, { "id": 63, "seek": 19266, "start": 192.66, "end": 195.57999999999998, "text": "当然为了真实介入工业级开发流程", "tokens": [ 50365, 40486, 13992, 2289, 6303, 24726, 30312, 14028, 23323, 940, 248, 16853, 100, 18937, 28926, 27854, 29649, 50511 ], "temperature": 0, "avg_logprob": -0.09798635967037221, "compression_ratio": 1.2621082621082622, "no_speech_prob": 1.7464165530389764e-11, "source": "whisper" }, { "id": 64, "seek": 19266, "start": 195.57999999999998, "end": 198.82, "text": "我们还需要给Deepseek加Kodex这套技术体系", "tokens": [ 50511, 15003, 14852, 35748, 23197, 11089, 595, 405, 916, 9990, 42, 1429, 87, 5562, 1881, 245, 32502, 1474, 107, 29485, 25368, 50673 ], "temperature": 0, "avg_logprob": -0.09798635967037221, "compression_ratio": 1.2621082621082622, "no_speech_prob": 1.7464165530389764e-11, "source": "whisper" }, { "id": 65, "seek": 19266, "start": 198.82, "end": 200.06, "text": "接上视觉能力", "tokens": [ 50673, 14468, 5708, 40656, 24447, 8225, 13486, 50735 ], "temperature": 0, "avg_logprob": -0.09798635967037221, "compression_ratio": 1.2621082621082622, "no_speech_prob": 1.7464165530389764e-11, "source": "whisper" }, { "id": 66, "seek": 19266, "start": 200.06, "end": 203.28, "text": "才能够在真实开发过程中实现开发成果的", "tokens": [ 50735, 18888, 8225, 1787, 253, 3581, 6303, 24726, 18937, 28926, 16866, 29649, 5975, 24726, 20204, 18937, 28926, 11336, 9319, 1546, 50896 ], "temperature": 0, "avg_logprob": -0.09798635967037221, "compression_ratio": 1.2621082621082622, "no_speech_prob": 1.7464165530389764e-11, "source": "whisper" }, { "id": 67, "seek": 19266, "start": 203.28, "end": 205.78, "text": "视觉自我审核外加审美配合对齐", "tokens": [ 50896, 40656, 24447, 9722, 1654, 2415, 94, 8987, 116, 12022, 9990, 2415, 94, 9175, 38846, 14245, 8713, 44628, 238, 51021 ], "temperature": 0, "avg_logprob": -0.09798635967037221, "compression_ratio": 1.2621082621082622, "no_speech_prob": 1.7464165530389764e-11, "source": "whisper" }, { "id": 68, "seek": 19266, "start": 205.78, "end": 206.6, "text": "简单来说", "tokens": [ 51021, 11249, 222, 47446, 6912, 8090, 51062 ], "temperature": 0, "avg_logprob": -0.09798635967037221, "compression_ratio": 1.2621082621082622, "no_speech_prob": 1.7464165530389764e-11, "source": "whisper" }, { "id": 69, "seek": 19266, "start": 206.6, "end": 208.98, "text": "就是让agent通过看到自己的开发成果", "tokens": [ 51062, 5620, 33650, 559, 317, 19550, 16866, 18032, 17645, 1546, 18937, 28926, 11336, 9319, 51181 ], "temperature": 0, "avg_logprob": -0.09798635967037221, "compression_ratio": 1.2621082621082622, "no_speech_prob": 1.7464165530389764e-11, "source": "whisper" }, { "id": 70, "seek": 19266, "start": 208.98, "end": 210.85999999999999, "text": "来提升开发效果", "tokens": [ 51181, 6912, 20949, 41670, 18937, 28926, 43076, 9319, 51275 ], "temperature": 0, "avg_logprob": -0.09798635967037221, "compression_ratio": 1.2621082621082622, "no_speech_prob": 1.7464165530389764e-11, "source": "whisper" }, { "id": 71, "seek": 19266, "start": 210.85999999999999, "end": 213.66, "text": "国外科中将采用1013.8作为视觉工具", "tokens": [ 51275, 16086, 12022, 42091, 5975, 45456, 5873, 229, 9254, 3279, 7668, 13, 23, 11914, 13992, 40656, 24447, 23323, 39806, 51415 ], "temperature": 0, "avg_logprob": -0.09798635967037221, "compression_ratio": 1.2621082621082622, "no_speech_prob": 1.7464165530389764e-11, "source": "whisper" }, { "id": 72, "seek": 19266, "start": 213.66, "end": 216.12, "text": "并借助MCP来提供触发接口", "tokens": [ 51415, 3509, 114, 2122, 253, 37618, 44, 20049, 6912, 20949, 3254, 249, 6758, 99, 28926, 14468, 18144, 51538 ], "temperature": 0, "avg_logprob": -0.09798635967037221, "compression_ratio": 1.2621082621082622, "no_speech_prob": 1.7464165530389764e-11, "source": "whisper" }, { "id": 73, "seek": 19266, "start": 216.12, "end": 217.78, "text": "完整架构如图所示", "tokens": [ 51538, 14128, 27662, 7360, 114, 7360, 226, 8238, 3919, 122, 5966, 25696, 51621 ], "temperature": 0, "avg_logprob": -0.09798635967037221, "compression_ratio": 1.2621082621082622, "no_speech_prob": 1.7464165530389764e-11, "source": "whisper" }, { "id": 74, "seek": 19266, "start": 217.78, "end": 221.54, "text": "而Deepseek V4也只有在接入视觉能力之后", "tokens": [ 51621, 11070, 11089, 595, 405, 916, 691, 19, 6404, 35244, 3581, 14468, 14028, 40656, 24447, 8225, 13486, 9574, 13547, 51809 ], "temperature": 0, "avg_logprob": -0.09798635967037221, "compression_ratio": 1.2621082621082622, "no_speech_prob": 1.7464165530389764e-11, "source": "whisper" }, { "id": 75, "seek": 22154, "start": 221.54, "end": 224.84, "text": "才能够实现如图所示的非常惊艳的开发效果", "tokens": [ 50365, 18888, 8225, 1787, 253, 24726, 20204, 8238, 3919, 122, 5966, 25696, 1546, 14392, 4799, 232, 12531, 111, 1546, 18937, 28926, 43076, 9319, 50530 ], "temperature": 0, "avg_logprob": -0.07836801438104539, "compression_ratio": 1.3590504451038576, "no_speech_prob": 1.9456021863040185e-11, "source": "whisper" }, { "id": 76, "seek": 22154, "start": 224.84, "end": 227.84, "text": "而在完成了一系列的基础准备工作之后", "tokens": [ 50530, 11070, 3581, 41509, 2289, 2257, 25368, 43338, 1546, 26008, 38114, 222, 6336, 228, 1787, 229, 41315, 9574, 13547, 50680 ], "temperature": 0, "avg_logprob": -0.07836801438104539, "compression_ratio": 1.3590504451038576, "no_speech_prob": 1.9456021863040185e-11, "source": "whisper" }, { "id": 77, "seek": 22154, "start": 227.84, "end": 228.84, "text": "工坏课的下半场", "tokens": [ 50680, 23323, 14872, 237, 5233, 122, 1546, 4438, 30018, 50255, 50730 ], "temperature": 0, "avg_logprob": -0.07836801438104539, "compression_ratio": 1.3590504451038576, "no_speech_prob": 1.9456021863040185e-11, "source": "whisper" }, { "id": 78, "seek": 22154, "start": 228.84, "end": 231.6, "text": "就让我们进入工业级agent的开发实战阶段", "tokens": [ 50730, 3111, 33650, 15003, 36700, 14028, 23323, 940, 248, 16853, 100, 559, 317, 1546, 18937, 28926, 24726, 1486, 246, 10034, 114, 28427, 50868 ], "temperature": 0, "avg_logprob": -0.07836801438104539, "compression_ratio": 1.3590504451038576, "no_speech_prob": 1.9456021863040185e-11, "source": "whisper" }, { "id": 79, "seek": 22154, "start": 231.6, "end": 232.62, "text": "本期工坏课", "tokens": [ 50868, 8802, 16786, 23323, 14872, 237, 5233, 122, 50919 ], "temperature": 0, "avg_logprob": -0.07836801438104539, "compression_ratio": 1.3590504451038576, "no_speech_prob": 1.9456021863040185e-11, "source": "whisper" }, { "id": 80, "seek": 22154, "start": 232.62, "end": 236.07999999999998, "text": "我们将同样选取目前最热门的agent的应用方向", "tokens": [ 50919, 15003, 45456, 13089, 14496, 2215, 231, 29436, 39004, 8661, 23661, 255, 8259, 101, 1546, 559, 317, 1546, 44297, 9254, 9249, 24282, 51092 ], "temperature": 0, "avg_logprob": -0.07836801438104539, "compression_ratio": 1.3590504451038576, "no_speech_prob": 1.9456021863040185e-11, "source": "whisper" }, { "id": 81, "seek": 22154, "start": 236.07999999999998, "end": 237.18, "text": "AI数据分析领域", "tokens": [ 51092, 48698, 33188, 26075, 106, 6627, 7360, 238, 12501, 228, 16262, 253, 51147 ], "temperature": 0, "avg_logprob": -0.07836801438104539, "compression_ratio": 1.3590504451038576, "no_speech_prob": 1.9456021863040185e-11, "source": "whisper" }, { "id": 82, "seek": 22154, "start": 237.18, "end": 237.84, "text": "详细介绍", "tokens": [ 51147, 5233, 99, 10115, 228, 30312, 10115, 235, 51180 ], "temperature": 0, "avg_logprob": -0.07836801438104539, "compression_ratio": 1.3590504451038576, "no_speech_prob": 1.9456021863040185e-11, "source": "whisper" }, { "id": 83, "seek": 22154, "start": 237.84, "end": 241.42, "text": "如何搭建一整套工业级可用的AI数据分析系统", "tokens": [ 51180, 43526, 22220, 255, 34157, 2257, 27662, 1881, 245, 23323, 940, 248, 16853, 100, 4429, 9254, 1546, 48698, 33188, 26075, 106, 6627, 7360, 238, 25368, 10115, 253, 51359 ], "temperature": 0, "avg_logprob": -0.07836801438104539, "compression_ratio": 1.3590504451038576, "no_speech_prob": 1.9456021863040185e-11, "source": "whisper" }, { "id": 84, "seek": 22154, "start": 241.42, "end": 243.54, "text": "其实经历了这么多年的发展", "tokens": [ 51359, 41646, 30276, 5014, 228, 2289, 31889, 6392, 34128, 28926, 43491, 51465 ], "temperature": 0, "avg_logprob": -0.07836801438104539, "compression_ratio": 1.3590504451038576, "no_speech_prob": 1.9456021863040185e-11, "source": "whisper" }, { "id": 85, "seek": 22154, "start": 243.54, "end": 244.39999999999998, "text": "工业场景下", "tokens": [ 51465, 23323, 940, 248, 50255, 50218, 4438, 51508 ], "temperature": 0, "avg_logprob": -0.07836801438104539, "compression_ratio": 1.3590504451038576, "no_speech_prob": 1.9456021863040185e-11, "source": "whisper" }, { "id": 86, "seek": 22154, "start": 244.39999999999998, "end": 245.7, "text": "对于数据分析智能体", "tokens": [ 51508, 8713, 37732, 33188, 26075, 106, 6627, 7360, 238, 5094, 118, 8225, 29485, 51573 ], "temperature": 0, "avg_logprob": -0.07836801438104539, "compression_ratio": 1.3590504451038576, "no_speech_prob": 1.9456021863040185e-11, "source": "whisper" }, { "id": 87, "seek": 22154, "start": 245.7, "end": 247.57999999999998, "text": "早就不再是简单的提个数", "tokens": [ 51573, 21176, 3111, 1960, 8623, 1541, 11249, 222, 47446, 1546, 20949, 7549, 33188, 51667 ], "temperature": 0, "avg_logprob": -0.07836801438104539, "compression_ratio": 1.3590504451038576, "no_speech_prob": 1.9456021863040185e-11, "source": "whisper" }, { "id": 88, "seek": 24758, "start": 247.58, "end": 249.68, "text": "绘的图进行数据洞察分析", "tokens": [ 50365, 10115, 246, 1546, 3919, 122, 36700, 8082, 33188, 26075, 106, 12136, 252, 47550, 6627, 7360, 238, 50470 ], "temperature": 0, "avg_logprob": -0.0972657994839108, "compression_ratio": 1.3700564971751412, "no_speech_prob": 1.4276719563499629e-11, "source": "whisper" }, { "id": 89, "seek": 24758, "start": 249.68, "end": 251.84, "text": "得出几条结论这么简单的需求了", "tokens": [ 50470, 5916, 7781, 6336, 254, 48837, 45641, 7422, 118, 31889, 11249, 222, 47446, 1546, 32535, 32718, 2289, 50578 ], "temperature": 0, "avg_logprob": -0.0972657994839108, "compression_ratio": 1.3700564971751412, "no_speech_prob": 1.4276719563499629e-11, "source": "whisper" }, { "id": 90, "seek": 24758, "start": 251.84, "end": 254.38000000000002, "text": "而是首先要能接入各种数据员", "tokens": [ 50578, 11070, 1541, 36490, 4275, 8225, 14468, 14028, 17516, 39810, 33188, 26075, 106, 3606, 246, 50705 ], "temperature": 0, "avg_logprob": -0.0972657994839108, "compression_ratio": 1.3700564971751412, "no_speech_prob": 1.4276719563499629e-11, "source": "whisper" }, { "id": 91, "seek": 24758, "start": 254.38000000000002, "end": 257.76, "text": "并至少支持百万级以上数据集的管理跟读取", "tokens": [ 50705, 3509, 114, 20844, 15686, 35488, 31906, 23570, 16853, 100, 29497, 33188, 26075, 106, 26020, 1546, 23131, 13876, 9678, 5233, 119, 29436, 50874 ], "temperature": 0, "avg_logprob": -0.0972657994839108, "compression_ratio": 1.3700564971751412, "no_speech_prob": 1.4276719563499629e-11, "source": "whisper" }, { "id": 92, "seek": 24758, "start": 257.76, "end": 259.90000000000003, "text": "同时还要支持数据信息", "tokens": [ 50874, 13089, 15729, 14852, 4275, 35488, 33188, 26075, 106, 17665, 26460, 50981 ], "temperature": 0, "avg_logprob": -0.0972657994839108, "compression_ratio": 1.3700564971751412, "no_speech_prob": 1.4276719563499629e-11, "source": "whisper" }, { "id": 93, "seek": 24758, "start": 259.90000000000003, "end": 260.82, "text": "可视化分析", "tokens": [ 50981, 4429, 40656, 23756, 6627, 7360, 238, 51027 ], "temperature": 0, "avg_logprob": -0.0972657994839108, "compression_ratio": 1.3700564971751412, "no_speech_prob": 1.4276719563499629e-11, "source": "whisper" }, { "id": 94, "seek": 24758, "start": 260.82, "end": 261.68, "text": "数据洞察", "tokens": [ 51027, 33188, 26075, 106, 12136, 252, 47550, 51070 ], "temperature": 0, "avg_logprob": -0.0972657994839108, "compression_ratio": 1.3700564971751412, "no_speech_prob": 1.4276719563499629e-11, "source": "whisper" }, { "id": 95, "seek": 24758, "start": 261.68, "end": 262.44, "text": "特征工程", "tokens": [ 51070, 17682, 2172, 223, 23323, 29649, 51108 ], "temperature": 0, "avg_logprob": -0.0972657994839108, "compression_ratio": 1.3700564971751412, "no_speech_prob": 1.4276719563499629e-11, "source": "whisper" }, { "id": 96, "seek": 24758, "start": 262.44, "end": 263.2, "text": "模型融合", "tokens": [ 51108, 41908, 39823, 43772, 235, 14245, 51146 ], "temperature": 0, "avg_logprob": -0.0972657994839108, "compression_ratio": 1.3700564971751412, "no_speech_prob": 1.4276719563499629e-11, "source": "whisper" }, { "id": 97, "seek": 24758, "start": 263.2, "end": 265.58000000000004, "text": "机器学习建模分析等完整链路", "tokens": [ 51146, 37960, 34386, 29618, 2930, 254, 34157, 41908, 6627, 7360, 238, 10187, 14128, 27662, 165, 241, 122, 24658, 51265 ], "temperature": 0, "avg_logprob": -0.0972657994839108, "compression_ratio": 1.3700564971751412, "no_speech_prob": 1.4276719563499629e-11, "source": "whisper" }, { "id": 98, "seek": 24758, "start": 265.58000000000004, "end": 267.04, "text": "以及还要能一键创建", "tokens": [ 51265, 40282, 14852, 4275, 8225, 2257, 23049, 106, 2437, 249, 34157, 51338 ], "temperature": 0, "avg_logprob": -0.0972657994839108, "compression_ratio": 1.3700564971751412, "no_speech_prob": 1.4276719563499629e-11, "source": "whisper" }, { "id": 99, "seek": 24758, "start": 267.04, "end": 269.94, "text": "满足具体业务格式和指标要求的数据分析报告", "tokens": [ 51338, 15868, 94, 37236, 39806, 29485, 940, 248, 5087, 94, 30921, 27584, 12565, 25922, 162, 3921, 4275, 32718, 1546, 33188, 26075, 106, 6627, 7360, 238, 49817, 16846, 51483 ], "temperature": 0, "avg_logprob": -0.0972657994839108, "compression_ratio": 1.3700564971751412, "no_speech_prob": 1.4276719563499629e-11, "source": "whisper" }, { "id": 100, "seek": 24758, "start": 269.94, "end": 271.92, "text": "此外还要有SKILLS功能系统", "tokens": [ 51483, 17947, 12022, 14852, 4275, 2412, 50, 42, 8621, 50, 32311, 8225, 25368, 10115, 253, 51582 ], "temperature": 0, "avg_logprob": -0.0972657994839108, "compression_ratio": 1.3700564971751412, "no_speech_prob": 1.4276719563499629e-11, "source": "whisper" }, { "id": 101, "seek": 24758, "start": 271.92, "end": 274.8, "text": "支持上传导入跟使用自定义的SKILL", "tokens": [ 51582, 35488, 5708, 7384, 254, 4510, 120, 14028, 9678, 22982, 9254, 9722, 12088, 2930, 231, 1546, 50, 42, 8621, 51726 ], "temperature": 0, "avg_logprob": -0.0972657994839108, "compression_ratio": 1.3700564971751412, "no_speech_prob": 1.4276719563499629e-11, "source": "whisper" }, { "id": 102, "seek": 27480, "start": 274.8, "end": 277.02000000000004, "text": "从而更好的进行功能拓展", "tokens": [ 50365, 35630, 11070, 19002, 20715, 36700, 8082, 32311, 8225, 6852, 241, 43491, 50476 ], "temperature": 0, "avg_logprob": -0.0997902182645576, "compression_ratio": 1.3461538461538463, "no_speech_prob": 1.5368756614564383e-11, "source": "whisper" }, { "id": 103, "seek": 27480, "start": 277.02000000000004, "end": 278.42, "text": "而这也是本期公开课中", "tokens": [ 50476, 11070, 5562, 22021, 8802, 16786, 13545, 18937, 5233, 122, 5975, 50546 ], "temperature": 0, "avg_logprob": -0.0997902182645576, "compression_ratio": 1.3461538461538463, "no_speech_prob": 1.5368756614564383e-11, "source": "whisper" }, { "id": 104, "seek": 27480, "start": 278.42, "end": 280.46000000000004, "text": "即将为大家深入系统讲解的", "tokens": [ 50546, 39127, 45456, 13992, 6868, 24043, 14028, 25368, 10115, 253, 39255, 17278, 1546, 50648 ], "temperature": 0, "avg_logprob": -0.0997902182645576, "compression_ratio": 1.3461538461538463, "no_speech_prob": 1.5368756614564383e-11, "source": "whisper" }, { "id": 105, "seek": 27480, "start": 280.46000000000004, "end": 282.3, "text": "DS4 AI数据分析系统", "tokens": [ 50648, 11844, 19, 7318, 33188, 26075, 106, 6627, 7360, 238, 25368, 10115, 253, 50740 ], "temperature": 0, "avg_logprob": -0.0997902182645576, "compression_ratio": 1.3461538461538463, "no_speech_prob": 1.5368756614564383e-11, "source": "whisper" }, { "id": 106, "seek": 27480, "start": 282.3, "end": 283.44, "text": "作为教学项目", "tokens": [ 50740, 11914, 13992, 21936, 29618, 10178, 117, 11386, 50797 ], "temperature": 0, "avg_logprob": -0.0997902182645576, "compression_ratio": 1.3461538461538463, "no_speech_prob": 1.5368756614564383e-11, "source": "whisper" }, { "id": 107, "seek": 27480, "start": 283.44, "end": 284.2, "text": "我会从头", "tokens": [ 50797, 1654, 12949, 35630, 39862, 50835 ], "temperature": 0, "avg_logprob": -0.0997902182645576, "compression_ratio": 1.3461538461538463, "no_speech_prob": 1.5368756614564383e-11, "source": "whisper" }, { "id": 108, "seek": 27480, "start": 284.2, "end": 285.96000000000004, "text": "为大家详细介绍AI数据分析的", "tokens": [ 50835, 13992, 6868, 5233, 99, 10115, 228, 30312, 10115, 235, 48698, 33188, 26075, 106, 6627, 7360, 238, 1546, 50923 ], "temperature": 0, "avg_logprob": -0.0997902182645576, "compression_ratio": 1.3461538461538463, "no_speech_prob": 1.5368756614564383e-11, "source": "whisper" }, { "id": 109, "seek": 27480, "start": 285.96000000000004, "end": 287.08000000000004, "text": "基本业务背景", "tokens": [ 50923, 37946, 940, 248, 5087, 94, 46329, 50218, 50979 ], "temperature": 0, "avg_logprob": -0.0997902182645576, "compression_ratio": 1.3461538461538463, "no_speech_prob": 1.5368756614564383e-11, "source": "whisper" }, { "id": 110, "seek": 27480, "start": 287.08000000000004, "end": 288.54, "text": "核心项目架构", "tokens": [ 50979, 8987, 116, 7945, 10178, 117, 11386, 7360, 114, 7360, 226, 51052 ], "temperature": 0, "avg_logprob": -0.0997902182645576, "compression_ratio": 1.3461538461538463, "no_speech_prob": 1.5368756614564383e-11, "source": "whisper" }, { "id": 111, "seek": 27480, "start": 288.54, "end": 290.76, "text": "以及完整的技术选进方案", "tokens": [ 51052, 40282, 14128, 27662, 1546, 32502, 1474, 107, 2215, 231, 36700, 9249, 28899, 51163 ], "temperature": 0, "avg_logprob": -0.0997902182645576, "compression_ratio": 1.3461538461538463, "no_speech_prob": 1.5368756614564383e-11, "source": "whisper" }, { "id": 112, "seek": 27480, "start": 290.76, "end": 293.7, "text": "并借助Codex一步步完成项目开发", "tokens": [ 51163, 3509, 114, 2122, 253, 37618, 34, 1429, 87, 2257, 31429, 31429, 41509, 10178, 117, 11386, 18937, 28926, 51310 ], "temperature": 0, "avg_logprob": -0.0997902182645576, "compression_ratio": 1.3461538461538463, "no_speech_prob": 1.5368756614564383e-11, "source": "whisper" }, { "id": 113, "seek": 27480, "start": 293.7, "end": 294.34000000000003, "text": "此外", "tokens": [ 51310, 17947, 12022, 51342 ], "temperature": 0, "avg_logprob": -0.0997902182645576, "compression_ratio": 1.3461538461538463, "no_speech_prob": 1.5368756614564383e-11, "source": "whisper" }, { "id": 114, "seek": 27480, "start": 294.34000000000003, "end": 295.8, "text": "为了便于同学们之后", "tokens": [ 51342, 13992, 2289, 27364, 37732, 13089, 29618, 9497, 9574, 13547, 51415 ], "temperature": 0, "avg_logprob": -0.0997902182645576, "compression_ratio": 1.3461538461538463, "no_speech_prob": 1.5368756614564383e-11, "source": "whisper" }, { "id": 115, "seek": 27480, "start": 295.8, "end": 297.56, "text": "自己也能熟练操作Codex", "tokens": [ 51415, 17645, 6404, 8225, 21805, 253, 10115, 225, 13167, 235, 11914, 34, 1429, 87, 51503 ], "temperature": 0, "avg_logprob": -0.0997902182645576, "compression_ratio": 1.3461538461538463, "no_speech_prob": 1.5368756614564383e-11, "source": "whisper" }, { "id": 116, "seek": 27480, "start": 297.56, "end": 298.76, "text": "完成自己的项目开发", "tokens": [ 51503, 41509, 17645, 1546, 10178, 117, 11386, 18937, 28926, 51563 ], "temperature": 0, "avg_logprob": -0.0997902182645576, "compression_ratio": 1.3461538461538463, "no_speech_prob": 1.5368756614564383e-11, "source": "whisper" }, { "id": 117, "seek": 27480, "start": 298.76, "end": 299.24, "text": "公开课中", "tokens": [ 51563, 13545, 18937, 5233, 122, 5975, 51587 ], "temperature": 0, "avg_logprob": -0.0997902182645576, "compression_ratio": 1.3461538461538463, "no_speech_prob": 1.5368756614564383e-11, "source": "whisper" }, { "id": 118, "seek": 27480, "start": 299.24, "end": 300.84000000000003, "text": "还特地为大家准备了一份", "tokens": [ 51587, 14852, 17682, 10928, 13992, 6868, 6336, 228, 1787, 229, 2289, 2257, 36266, 51667 ], "temperature": 0, "avg_logprob": -0.0997902182645576, "compression_ratio": 1.3461538461538463, "no_speech_prob": 1.5368756614564383e-11, "source": "whisper" }, { "id": 119, "seek": 27480, "start": 300.84000000000003, "end": 302.56, "text": "完整的DeepSeek V4", "tokens": [ 51667, 14128, 27662, 1546, 11089, 595, 10637, 916, 691, 19, 51753 ], "temperature": 0, "avg_logprob": -0.0997902182645576, "compression_ratio": 1.3461538461538463, "no_speech_prob": 1.5368756614564383e-11, "source": "whisper" }, { "id": 120, "seek": 30256, "start": 302.56, "end": 304.78000000000003, "text": "技术体系LMWiki知识库", "tokens": [ 50365, 32502, 1474, 107, 29485, 25368, 43, 44, 54, 9850, 6498, 5233, 228, 6346, 241, 50476 ], "temperature": 0, "avg_logprob": -0.16981897700018217, "compression_ratio": 1.1851851851851851, "no_speech_prob": 1.587599496311043e-11, "source": "whisper" }, { "id": 121, "seek": 30256, "start": 304.78000000000003, "end": 308.46, "text": "以及10项DeepSeek V4开发专属SKILLS", "tokens": [ 50476, 40282, 3279, 10178, 117, 11089, 595, 10637, 916, 691, 19, 18937, 28926, 940, 241, 9636, 252, 50, 42, 8621, 50, 50660 ], "temperature": 0, "avg_logprob": -0.16981897700018217, "compression_ratio": 1.1851851851851851, "no_speech_prob": 1.587599496311043e-11, "source": "whisper" }, { "id": 122, "seek": 30256, "start": 308.46, "end": 310.1, "text": "目前全能公开课的课件", "tokens": [ 50660, 39004, 11319, 8225, 13545, 18937, 5233, 122, 1546, 5233, 122, 20485, 50742 ], "temperature": 0, "avg_logprob": -0.16981897700018217, "compression_ratio": 1.1851851851851851, "no_speech_prob": 1.587599496311043e-11, "source": "whisper" }, { "id": 123, "seek": 30256, "start": 310.1, "end": 311.42, "text": "脚本项目提示词", "tokens": [ 50742, 27067, 248, 8802, 10178, 117, 11386, 20949, 25696, 5233, 235, 50808 ], "temperature": 0, "avg_logprob": -0.16981897700018217, "compression_ratio": 1.1851851851851851, "no_speech_prob": 1.587599496311043e-11, "source": "whisper" }, { "id": 124, "seek": 30256, "start": 311.42, "end": 312.8, "text": "知识库和SKILLS的", "tokens": [ 50808, 6498, 5233, 228, 6346, 241, 12565, 50, 42, 8621, 50, 1546, 50877 ], "temperature": 0, "avg_logprob": -0.16981897700018217, "compression_ratio": 1.1851851851851851, "no_speech_prob": 1.587599496311043e-11, "source": "whisper" }, { "id": 125, "seek": 30256, "start": 312.8, "end": 314.78000000000003, "text": "都已经上限至复返AI社区", "tokens": [ 50877, 7182, 49161, 5708, 43446, 20844, 1787, 235, 3316, 242, 48698, 27658, 9937, 118, 50976 ], "temperature": 0, "avg_logprob": -0.16981897700018217, "compression_ratio": 1.1851851851851851, "no_speech_prob": 1.587599496311043e-11, "source": "whisper" }, { "id": 126, "seek": 30256, "start": 314.78000000000003, "end": 315.88, "text": "加扫标评商的奥运码", "tokens": [ 50976, 9990, 3416, 104, 162, 3921, 5233, 226, 45581, 1546, 1881, 98, 3316, 238, 23230, 223, 51031 ], "temperature": 0, "avg_logprob": -0.16981897700018217, "compression_ratio": 1.1851851851851851, "no_speech_prob": 1.587599496311043e-11, "source": "whisper" }, { "id": 127, "seek": 30256, "start": 315.88, "end": 317.02, "text": "即可免费领取", "tokens": [ 51031, 39127, 4429, 2347, 235, 18464, 117, 12501, 228, 29436, 51088 ], "temperature": 0, "avg_logprob": -0.16981897700018217, "compression_ratio": 1.1851851851851851, "no_speech_prob": 1.587599496311043e-11, "source": "whisper" }, { "id": 128, "seek": 30256, "start": 317.02, "end": 319.04, "text": "2016大模型Agent的技术发展", "tokens": [ 51088, 41103, 3582, 41908, 39823, 32, 6930, 1546, 32502, 1474, 107, 28926, 43491, 51189 ], "temperature": 0, "avg_logprob": -0.16981897700018217, "compression_ratio": 1.1851851851851851, "no_speech_prob": 1.587599496311043e-11, "source": "whisper" }, { "id": 129, "seek": 30256, "start": 319.04, "end": 319.94, "text": "是都非常迅猛", "tokens": [ 51189, 1541, 7182, 14392, 3316, 227, 25266, 249, 51234 ], "temperature": 0, "avg_logprob": -0.16981897700018217, "compression_ratio": 1.1851851851851851, "no_speech_prob": 1.587599496311043e-11, "source": "whisper" }, { "id": 130, "seek": 30256, "start": 319.94, "end": 321.16, "text": "我也将持续为大家提供", "tokens": [ 51234, 27761, 45456, 17694, 10115, 255, 13992, 6868, 20949, 3254, 249, 51295 ], "temperature": 0, "avg_logprob": -0.16981897700018217, "compression_ratio": 1.1851851851851851, "no_speech_prob": 1.587599496311043e-11, "source": "whisper" }, { "id": 131, "seek": 30256, "start": 321.16, "end": 322.86, "text": "最前沿实用的技术教学", "tokens": [ 51295, 8661, 8945, 3308, 123, 24726, 9254, 1546, 32502, 1474, 107, 21936, 29618, 51380 ], "temperature": 0, "avg_logprob": -0.16981897700018217, "compression_ratio": 1.1851851851851851, "no_speech_prob": 1.587599496311043e-11, "source": "whisper" }, { "id": 132, "seek": 30256, "start": 322.86, "end": 324.6, "text": "感谢大家的关注和三连支持", "tokens": [ 51380, 9709, 11340, 6868, 1546, 28053, 26432, 12565, 10960, 3316, 252, 35488, 51467 ], "temperature": 0, "avg_logprob": -0.16981897700018217, "compression_ratio": 1.1851851851851851, "no_speech_prob": 1.587599496311043e-11, "source": "whisper" }, { "id": 133, "seek": 30256, "start": 324.6, "end": 325.12, "text": "公开课", "tokens": [ 51467, 13545, 18937, 5233, 122, 51493 ], "temperature": 0, "avg_logprob": -0.16981897700018217, "compression_ratio": 1.1851851851851851, "no_speech_prob": 1.587599496311043e-11, "source": "whisper" }, { "id": 134, "seek": 30256, "start": 325.12, "end": 326.0, "text": "马上开始", "tokens": [ 51493, 49970, 5708, 45213, 51537 ], "temperature": 0, "avg_logprob": -0.16981897700018217, "compression_ratio": 1.1851851851851851, "no_speech_prob": 1.587599496311043e-11, "source": "whisper" } ], "transcript_selection": { "schema": "videonote.transcript_selection.v1", "status": "SELECTED", "requested_language": null, "duration": 326.33325, "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": null, "matched": false } }, "whisper_profile": { "segment_count": 135, "multilingual": false, "first_start": 0.0, "final_end": 326.0, "duration_coverage": 0.998979, "maximum_gap_seconds": 0.0, "inverted_count": 0, "language": { "requested": null, "matched": true, "latin": 528, "han": 1544, "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": [] }, "filled_source_profile": null, "whisper_duration_guard": { "media_duration": 326.333, "rejected_segment_count": 0, "rejected_segments": [] } } }