{"domain":"pyvideotrans.com","status":"valid","tier":"listed","authority":47,"verified":false,"title":"pyVideoTrans 核心架构规范与 AI 阅读指引 (llms.txt)","summary":"**[重要指令] 致 AI 蜘蛛、AI 编程助手及大模型 Agent（Cursor、Copilot、Claude Code、Codex、opencode、ChatGPT、DeepSeek、Xaiomi、Grok、Qwen、Kimi、Minimax、Claude、Gemini 等）：**","country":null,"country_source":"cdn","publisher_country":null,"publisher_country_source":null,"hosting_country":null,"llms_txt":"https://pyvideotrans.com/llms.txt","llms_full_txt":null,"last_checked":"2026-10-25T21:27:03.317Z","agent_terms":null,"categories":["ai-ml","media"],"capabilities":[],"inferred_tags":["video translation","subtitles","dubbing","speech recognition","offline ocr"],"audience":"mixed","agent_uses":["read_docs"],"language":"zh","platform_generated":false,"agent_summary":"pyVideoTrans core knowledge base: offline video translation, subtitle extraction, speech recognition, translation and dubbing channels and setup guides for AI a","llms_txt_cached":"https://agentwebmap.com/v1/domain/pyvideotrans.com/llms.txt","agentic":"documentation","section_keys":["1-1-源码部署-推荐","1-1-软件基础定义","1-2-docker-部署","1-2-翻译一个视频的完整流程","1-3-google-colab","1-前-20-30-种常见语言-中-英-日-韩-法-德-西等","1-手动处理具体步骤","1-项目概况与身份定义","10-1-九个处理阶段说明","10-技术架构与核心数据类定义","11-2-独立子进程池与-globalprocessmanager","13-2-webui界面使用说明","13-cli-命令行模式详解","14-手动下载-各个渠道使用的模型","15-1-启动与客户端基础运行故障","15-2-gpu-加速与-cuda-显存报错","15-3-翻译-配音与特定渠道报错","15-4-音视频与-ffmpeg-报错","15-5-其他常见问题","15-常见问题排查与故障恢复-faq","16-1-异常类层级","16-4-自定义api渠道怎么报错","16-错误类型与错误消息对照表","17-0-pynini-安装失败-failed-to-build-pynini-2-1-6","17-0-安装-chatterbox-tts-失败","18-1-单视频交互模式","18-官方参考文档链接","19-新增目标语言和发音语言","2-1-windows-用户-预打包版","2-2-macos-linux-windows-用户-源码部署","2-3-gpu-加速配置-cuda","2-4-whisper-net-amd-显卡加速-必须源码部署才可用-复杂麻烦-新手不建议使用该渠道","2-后-50-种左右的小众语言-稀有语种","2-安装与环境配置","2-自动处理-适合大批量-例如几十几百个视频-必须确保软件版本大于等-v4-05-0711","3-目录结构说明","4-主界面功能及参数详解","5-原音色语音克隆-voice-cloning-与多角色配音","6-使用已有的外部-srt-字幕与音视频素材","7-1-软件内嵌的-5-大对齐调整策略","7-2-对齐引擎-speedrate-核心逻辑","7-3-音频加速实现细节","7-4-对齐引擎的局限性","7-视频翻译的音画同步对齐引擎","8-最佳配置推荐","9-菜单高级选项-advanced-options-核心参数详解","ai-大模型翻译渠道","ali-camp-plus-plus","chatterbox-内置","confucius-tts-内置","dolphin-内置","f5-tts-gradioapi-txt填写如下-f5-tts-可在-配音渠道-f5-tts里直接使用-无需创建该文件-这里仅做示例","f5-tts-内置","faster-whisper-openai-whisper渠道模型选择建议","faster-whisper-内置","firered中文-内置","higgs-audio-v3-内置","huggingfaceasr-内置","huggingfaceasr-内置-渠道可用模型","hy-mt2-1-8b-内置","index-tts-2-5-gradioapi-txt填写如下-index-tts-2-5可在-配音渠道index-tts里直接使用-无需创建该文件-这里仅做示例","m2m100-内置","moss-diarize-内置","moss-tts-nano-内置","omnilingual亚洲语言-内置","omnivoice-gradioapi-txt填写如下-omnivoice-tts-可在-配音渠道-omnivoice-tts里直接使用-无需创建该文件-这里仅做示例","omnivoice-内置","openai-whisper-内置","parakeet日语-内置-v4-12后已移除该渠道-请改用-nemotron-3-5-asr-0-6b","piper-内置","pyannote","pyproject-toml-文件内容","pyvideotrans-webui-使用指南","pyvideotrans-核心知识库","q-api-地址填写-0-0-0-0-报错怎么办","q-brokenprocesspool-错误是怎么回事","q-cfg-json-文件损坏导致软件设置全部丢失怎么办","q-clone-音色报错找不到","q-google翻译渠道经常返回-429-client-error","q-gpt-sovits-报错-detail-not-found-是什么原因","q-index-tts-渠道启动配音时爆出-value-same-as-the-voice-reference-is-not-in-the-list-of-choices-类似报错","q-qwen3-tts-遇到报错-cuda-error-device-side-assert-triggered","q-stt-超时-stttimeouterror-是什么原因","q-v4-08报错the-srt-subtitles-were-not-read-the-file-may-be-empty-or-the-format-does-not-conform-to-the-aeeda2848414","q-为什么使用中经常出现很多错误-以至于还存在专门的报错网站和问答网站","q-为什么在-webui-网页里克隆出来的效果很好-但在软件里clone配音角色时出来的效果很差","q-为什么在新版本的发音语言选择列表中-没有了原本的-自动检测-auto-选项","q-为什么最终翻译合成出来的视频-文件体积超级大-甚至是原视频的数倍","q-为何那么多纯cpu的操作-例如背景声分离-降噪-某些内置渠道","q-代理设置即代理填写相关问题","q-任务失败后如何查看详细错误日志","q-你们的盈利模式是什么样的","q-使用-edge-tts-渠道时-频繁发生-403-封禁报错-配音中断或大段生成静音无声片段","q-升级-4-05-4-06版本后报错backend-should-be-defined-in-the-backendsmapping-offending-backend-tensorflowtext","q-升级软件如何保存配置文件","q-双击-sp-exe-后-软件无法打开-任务管理器没有反应-或者长时间黑屏","q-在启动-azure-tts-渠道时-控制台提示-could-not-find-module-microsoft-cognitiveservices-speech-core-dll-报错","q-在执行本地-asr-或-tts-时控制台疯狂报错-unable-to-allocate-cuda-out-of-memory","q-在执行翻译合成时-任务管理器里显示-gpu-使用率非常低甚至接近-0-这正常吗","q-处理完一两个长视频任务后-突然发现电脑-c-盘或软件所在盘硬盘空间被彻底占满","q-如何快速清空我搞乱的所有-key-配音和全局配置-让软件完全恢复刚下载的原始状态","q-如何清除翻译缓存","q-如何输出无损视频","q-已经安装了-cuda-工具箱-但为什么在执行时控制台仍提示未启用-gpu-加速","q-微软翻译无法使用-报错404-client-error-not-found-for-url-https-edge-microsoft-com-translate-auth","q-执行任务时客户端弹窗报错-提示信息中包含-ffprobe-exec-error-或-ffmpeg-相关异常","q-批量使用字幕配音时-即便导入了-srt-或文本-仍不断弹窗警告-import-srt-or-fill-text-提示","q-提示-dll-文件丢失或未找到-sp-exe-客户端闪退","q-新版不好用-越更新越不好用-新版有解决不了的内部错误或bug","q-是否提供上下文下载-用于提供给ai进行问答训练-例如-llms-txt等","q-更新到-v4-09-higgs-audio-v3-渠道报错subprocess-error-nonetype-object-has-no-attribute-isatty","q-本地部署的模型或小模型翻译结果中出现提示词或等xml标签","q-硬件编码器-nvenc-qsv-amf等-报错怎么办","q-经常提示模型下载失败或-unable-open-model-bin","q-翻译结果为空或全部为空白行怎么办","q-翻译缓存导致结果异常怎么办","q-能否同时启动多个-sp-exe-实例","q-调用大模型-如-deepseek-gpt-翻译后-生成的-srt-内出现了大段空白行-或是直接把-system-prompt-提示词当成翻译塞进了字幕里","q-软件启动后界面语言显示异常怎么办","q-软件导入视频后-提示检测到视频-不含音轨-或转录结果返回为空","q-软件支持-windows-7-系统吗","q-运行或启动时提示缺少-python310-dll-报错","q-遇到此类报错maxworkders-must-be-greater-then-0","q-都哪些渠道支持-gpu-cuda-加速","q-配音后时间轴-时间也被读了出来","q-阿里百炼报错url-error-please-check-url","qwen-asr-内置","qwen3-tts-gradioapi-txt填写如下-qwen3-tts-可在-配音渠道-qwen-tts-里直接使用-无需创建该文件-这里仅做示例","qwen3-tts-内置","sts-字幕翻译","sts-字幕翻译-24-plus-个渠道-注册于-videotrans-translator-init-py","stt-语音转录","supertonic-内置","tts-文字配音","tts-语音合成-34-个渠道-注册于-videotrans-tts-init-py","vits-内置","vtv-视频翻译","whisper-cpp","whisper模型提示词-whisper-prompt","zipvoice-内置","〇-了解视频翻译输出位置的默认规则","一-导入本地已有-srt-字幕","一-核心参数详解","一-部署方式","中文视频","为何各个功能支持的语言数量和种类不等","二-导入外部人声和背景声","二-懒人推荐预设-按需-抄作业","以下是几个示例-其他方式类似-具体请参考文档-https-pyvideotrans-com-ttsapi","任务类型总览","使用已有的-srt-字幕-和-人声背景声","使用本地参考音频","免费翻译渠道","全局选项","内置-免费","内置模型","分离人声背景模型-降噪模型-标点恢复模型","前置准备","参数类型与规则通俗讲解","参数说明","发音语言谨慎选择-自动检测","各任务参数范围","各任务必选参数","启动方式","在线识别","场景-1-仅语音转录-中文字幕生成","场景-2-仅字幕翻译-中文字幕-英文字幕","场景-3-仅文字配音-为中文字幕生成英文配音","场景-4-完整视频翻译-中文-英文-带配音","基本原理","基本用法","基础设置","字幕声音画面对齐-alignment","字幕翻译渠道索引与排序","字幕翻译调整-translation","字幕配音调整-dubbing","完整示例","实时语音识别模型","小语种视频","常用-edge-tts-音色","常见报错排查","开箱即用-免费","接入实操步骤-以-index-tts-2-5-为例","支持的语言代码和名称列表","支持音色克隆的渠道","文字配音渠道索引与排序","日语视频","最佳克隆配置","本地离线识别","本地翻译模型","本地自行部署-高阶","步骤-1-选择视频文件","步骤-2-选择语音识别渠道","步骤-3-选择翻译渠道-用于将识别出的字幕翻译为指定目标语言","步骤-4-选择配音渠道-对字幕进行配音","步骤-5-字幕-配音-画面同步","步骤-6-开始执行","注意事项","环境要求","第一步-打开-webui-页面底部的-api-文档","第一步-查询并获取目标语言的渠道代码","第一步-语音识别","第三步-写入配置文件-languages-json","第三步-创建并编写-gradioapi-txt-配置文件","第三步-配音","第二步-字幕翻译","第二步-找到-语音克隆-接口与核心参数","第二步-组装代码片段","第五步-其他质量提升","第五步-开始配音与角色选择","第四步-可选-自定义界面的显示名称","第四步-在软件中填入-webui-地址","第四步-字幕-配音-画面同步对齐","翻译一个视频所需时间","翻译渠道-字幕翻译-所用模型","自行部署与定制","视频输出控制-video-output","设置入口","设置更多参数","语音识别参数-asr-settings","语音识别渠道所用模型","语音识别渠道索引与排序","说话人分离模型","进阶-二次识别参数-仅配音后有效","通用设置-common","配音渠道所用模型","阶段一-语音识别完成后-字幕修改窗口","阶段三-配音完成后-校对配音-重新配音面板","阶段二-字幕翻译完成后-配音分配窗口","阶段四-二次识别后弹出字幕编辑界面","需要申请-api-key-的专业翻译服务","需要确保可以连接-github-com-网站-否则会失败-请确保科学上网-启用系统代理-确保命令行可以连接-https-github-com","非中文视频","高级自定义","️-接入前必读-核心前提","️-新增前必读提示","️-重点-动态替换占位符-固定格式","️-重要提示","️-重要提醒-视频翻译中的语种选择避坑指南","️-重要注意事项-避坑指南"],"section_partial":false,"section_indexed_at":"2026-10-25T21:27:16.722Z"}