[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-voice-cloning-app--Voice-Cloning-App":3,"tool-voice-cloning-app--Voice-Cloning-App":64},[4,17,27,35,43,56],{"id":5,"name":6,"github_repo":7,"description_zh":8,"stars":9,"difficulty_score":10,"last_commit_at":11,"category_tags":12,"status":16},3808,"stable-diffusion-webui","AUTOMATIC1111\u002Fstable-diffusion-webui","stable-diffusion-webui 是一个基于 Gradio 构建的网页版操作界面，旨在让用户能够轻松地在本地运行和使用强大的 Stable Diffusion 图像生成模型。它解决了原始模型依赖命令行、操作门槛高且功能分散的痛点，将复杂的 AI 绘图流程整合进一个直观易用的图形化平台。\n\n无论是希望快速上手的普通创作者、需要精细控制画面细节的设计师，还是想要深入探索模型潜力的开发者与研究人员，都能从中获益。其核心亮点在于极高的功能丰富度：不仅支持文生图、图生图、局部重绘（Inpainting）和外绘（Outpainting）等基础模式，还独创了注意力机制调整、提示词矩阵、负向提示词以及“高清修复”等高级功能。此外，它内置了 GFPGAN 和 CodeFormer 等人脸修复工具，支持多种神经网络放大算法，并允许用户通过插件系统无限扩展能力。即使是显存有限的设备，stable-diffusion-webui 也提供了相应的优化选项，让高质量的 AI 艺术创作变得触手可及。",162132,3,"2026-04-05T11:01:52",[13,14,15],"开发框架","图像","Agent","ready",{"id":18,"name":19,"github_repo":20,"description_zh":21,"stars":22,"difficulty_score":23,"last_commit_at":24,"category_tags":25,"status":16},1381,"everything-claude-code","affaan-m\u002Feverything-claude-code","everything-claude-code 是一套专为 AI 编程助手（如 Claude Code、Codex、Cursor 等）打造的高性能优化系统。它不仅仅是一组配置文件，而是一个经过长期实战打磨的完整框架，旨在解决 AI 代理在实际开发中面临的效率低下、记忆丢失、安全隐患及缺乏持续学习能力等核心痛点。\n\n通过引入技能模块化、直觉增强、记忆持久化机制以及内置的安全扫描功能，everything-claude-code 能显著提升 AI 在复杂任务中的表现，帮助开发者构建更稳定、更智能的生产级 AI 代理。其独特的“研究优先”开发理念和针对 Token 消耗的优化策略，使得模型响应更快、成本更低，同时有效防御潜在的攻击向量。\n\n这套工具特别适合软件开发者、AI 研究人员以及希望深度定制 AI 工作流的技术团队使用。无论您是在构建大型代码库，还是需要 AI 协助进行安全审计与自动化测试，everything-claude-code 都能提供强大的底层支持。作为一个曾荣获 Anthropic 黑客大奖的开源项目，它融合了多语言支持与丰富的实战钩子（hooks），让 AI 真正成长为懂上",138956,2,"2026-04-05T11:33:21",[13,15,26],"语言模型",{"id":28,"name":29,"github_repo":30,"description_zh":31,"stars":32,"difficulty_score":23,"last_commit_at":33,"category_tags":34,"status":16},2271,"ComfyUI","Comfy-Org\u002FComfyUI","ComfyUI 是一款功能强大且高度模块化的视觉 AI 引擎，专为设计和执行复杂的 Stable Diffusion 图像生成流程而打造。它摒弃了传统的代码编写模式，采用直观的节点式流程图界面，让用户通过连接不同的功能模块即可构建个性化的生成管线。\n\n这一设计巧妙解决了高级 AI 绘图工作流配置复杂、灵活性不足的痛点。用户无需具备编程背景，也能自由组合模型、调整参数并实时预览效果，轻松实现从基础文生图到多步骤高清修复等各类复杂任务。ComfyUI 拥有极佳的兼容性，不仅支持 Windows、macOS 和 Linux 全平台，还广泛适配 NVIDIA、AMD、Intel 及苹果 Silicon 等多种硬件架构，并率先支持 SDXL、Flux、SD3 等前沿模型。\n\n无论是希望深入探索算法潜力的研究人员和开发者，还是追求极致创作自由度的设计师与资深 AI 绘画爱好者，ComfyUI 都能提供强大的支持。其独特的模块化架构允许社区不断扩展新功能，使其成为当前最灵活、生态最丰富的开源扩散模型工具之一，帮助用户将创意高效转化为现实。",107662,"2026-04-03T11:11:01",[13,14,15],{"id":36,"name":37,"github_repo":38,"description_zh":39,"stars":40,"difficulty_score":23,"last_commit_at":41,"category_tags":42,"status":16},3704,"NextChat","ChatGPTNextWeb\u002FNextChat","NextChat 是一款轻量且极速的 AI 助手，旨在为用户提供流畅、跨平台的大模型交互体验。它完美解决了用户在多设备间切换时难以保持对话连续性，以及面对众多 AI 模型不知如何统一管理的痛点。无论是日常办公、学习辅助还是创意激发，NextChat 都能让用户随时随地通过网页、iOS、Android、Windows、MacOS 或 Linux 端无缝接入智能服务。\n\n这款工具非常适合普通用户、学生、职场人士以及需要私有化部署的企业团队使用。对于开发者而言，它也提供了便捷的自托管方案，支持一键部署到 Vercel 或 Zeabur 等平台。\n\nNextChat 的核心亮点在于其广泛的模型兼容性，原生支持 Claude、DeepSeek、GPT-4 及 Gemini Pro 等主流大模型，让用户在一个界面即可自由切换不同 AI 能力。此外，它还率先支持 MCP（Model Context Protocol）协议，增强了上下文处理能力。针对企业用户，NextChat 提供专业版解决方案，具备品牌定制、细粒度权限控制、内部知识库整合及安全审计等功能，满足公司对数据隐私和个性化管理的高标准要求。",87618,"2026-04-05T07:20:52",[13,26],{"id":44,"name":45,"github_repo":46,"description_zh":47,"stars":48,"difficulty_score":23,"last_commit_at":49,"category_tags":50,"status":16},2268,"ML-For-Beginners","microsoft\u002FML-For-Beginners","ML-For-Beginners 是由微软推出的一套系统化机器学习入门课程，旨在帮助零基础用户轻松掌握经典机器学习知识。这套课程将学习路径规划为 12 周，包含 26 节精炼课程和 52 道配套测验，内容涵盖从基础概念到实际应用的完整流程，有效解决了初学者面对庞大知识体系时无从下手、缺乏结构化指导的痛点。\n\n无论是希望转型的开发者、需要补充算法背景的研究人员，还是对人工智能充满好奇的普通爱好者，都能从中受益。课程不仅提供了清晰的理论讲解，还强调动手实践，让用户在循序渐进中建立扎实的技能基础。其独特的亮点在于强大的多语言支持，通过自动化机制提供了包括简体中文在内的 50 多种语言版本，极大地降低了全球不同背景用户的学习门槛。此外，项目采用开源协作模式，社区活跃且内容持续更新，确保学习者能获取前沿且准确的技术资讯。如果你正寻找一条清晰、友好且专业的机器学习入门之路，ML-For-Beginners 将是理想的起点。",84991,"2026-04-05T10:45:23",[14,51,52,53,15,54,26,13,55],"数据工具","视频","插件","其他","音频",{"id":57,"name":58,"github_repo":59,"description_zh":60,"stars":61,"difficulty_score":10,"last_commit_at":62,"category_tags":63,"status":16},3128,"ragflow","infiniflow\u002Fragflow","RAGFlow 是一款领先的开源检索增强生成（RAG）引擎，旨在为大语言模型构建更精准、可靠的上下文层。它巧妙地将前沿的 RAG 技术与智能体（Agent）能力相结合，不仅支持从各类文档中高效提取知识，还能让模型基于这些知识进行逻辑推理和任务执行。\n\n在大模型应用中，幻觉问题和知识滞后是常见痛点。RAGFlow 通过深度解析复杂文档结构（如表格、图表及混合排版），显著提升了信息检索的准确度，从而有效减少模型“胡编乱造”的现象，确保回答既有据可依又具备时效性。其内置的智能体机制更进一步，使系统不仅能回答问题，还能自主规划步骤解决复杂问题。\n\n这款工具特别适合开发者、企业技术团队以及 AI 研究人员使用。无论是希望快速搭建私有知识库问答系统，还是致力于探索大模型在垂直领域落地的创新者，都能从中受益。RAGFlow 提供了可视化的工作流编排界面和灵活的 API 接口，既降低了非算法背景用户的上手门槛，也满足了专业开发者对系统深度定制的需求。作为基于 Apache 2.0 协议开源的项目，它正成为连接通用大模型与行业专有知识之间的重要桥梁。",77062,"2026-04-04T04:44:48",[15,14,13,26,54],{"id":65,"github_repo":66,"name":67,"description_en":68,"description_zh":69,"ai_summary_zh":70,"readme_en":71,"readme_zh":72,"quickstart_zh":73,"use_case_zh":74,"hero_image_url":75,"owner_login":76,"owner_name":76,"owner_avatar_url":77,"owner_bio":78,"owner_company":78,"owner_location":78,"owner_email":78,"owner_twitter":78,"owner_website":78,"owner_url":79,"languages":80,"stars":105,"forks":106,"last_commit_at":107,"license":108,"difficulty_score":10,"env_os":109,"env_gpu":110,"env_ram":111,"env_deps":112,"category_tags":121,"github_topics":122,"view_count":129,"oss_zip_url":78,"oss_zip_packed_at":78,"status":16,"created_at":130,"updated_at":131,"faqs":132,"releases":161},527,"voice-cloning-app\u002FVoice-Cloning-App","Voice-Cloning-App","A Python\u002FPytorch app for easily synthesising human voices","Voice-Cloning-App 是一款基于 Python 和 PyTorch 构建的开源应用，致力于让用户能够轻松合成逼真的人类语音。在语音合成领域，传统的模型训练往往涉及复杂的代码编写和数据处理流程，Voice-Cloning-App 有效解决了这一痛点。它提供了一套完整的解决方案，从数据集自动生成到模型训练再到语音合成，大大降低了技术门槛。\n\nVoice-Cloning-App 特别适合开发者、AI 研究人员以及希望定制个性化语音的爱好者使用。即便没有深厚的编程背景，只要具备基本的电脑操作能力，也能利用其图形化界面完成工作。技术层面，Voice-Cloning-App 具有不少亮点：它支持自动构建数据集，兼容字幕文件和有声书；采用 HiFi-GAN 等先进模型提升音质；同时支持多 GPU 加速以及本地与远程训练的灵活切换。\n\n需要注意的是，运行 Voice-Cloning-App 通常需要 Windows 10 或 Ubuntu 系统，并建议配备 NVIDIA 显卡以获得最佳性能。总体而言，Voice-Cloning-App 为语音克隆技术提供了一个高效、易用的落地平台，让声音","Voice-Cloning-App 是一款基于 Python 和 PyTorch 构建的开源应用，致力于让用户能够轻松合成逼真的人类语音。在语音合成领域，传统的模型训练往往涉及复杂的代码编写和数据处理流程，Voice-Cloning-App 有效解决了这一痛点。它提供了一套完整的解决方案，从数据集自动生成到模型训练再到语音合成，大大降低了技术门槛。\n\nVoice-Cloning-App 特别适合开发者、AI 研究人员以及希望定制个性化语音的爱好者使用。即便没有深厚的编程背景，只要具备基本的电脑操作能力，也能利用其图形化界面完成工作。技术层面，Voice-Cloning-App 具有不少亮点：它支持自动构建数据集，兼容字幕文件和有声书；采用 HiFi-GAN 等先进模型提升音质；同时支持多 GPU 加速以及本地与远程训练的灵活切换。\n\n需要注意的是，运行 Voice-Cloning-App 通常需要 Windows 10 或 Ubuntu 系统，并建议配备 NVIDIA 显卡以获得最佳性能。总体而言，Voice-Cloning-App 为语音克隆技术提供了一个高效、易用的落地平台，让声音合成变得更加触手可及。","# Voice Cloning App\n[![CircleCI](https:\u002F\u002Fcircleci.com\u002Fgh\u002FBenAAndrew\u002FVoice-Cloning-App.svg?style=svg)](https:\u002F\u002Fcircleci.com\u002Fgh\u002FBenAAndrew\u002FVoice-Cloning-App)\n[![Discord](https:\u002F\u002Fimg.shields.io\u002Fdiscord\u002F833666557954883614.svg?style=flat-square)](https:\u002F\u002Fdiscord.gg\u002FwQd7zKCWxT)\n[![codecov](https:\u002F\u002Fcodecov.io\u002Fgh\u002FBenAAndrew\u002FVoice-Cloning-App\u002Fbranch\u002Fmain\u002Fgraph\u002Fbadge.svg?token=WC0LLZO3Z5)](https:\u002F\u002Fcodecov.io\u002Fgh\u002FBenAAndrew\u002FVoice-Cloning-App)\n[![comment](https:\u002F\u002Fcircleci.com\u002Fapi\u002Fv1.1\u002Fproject\u002Fgithub\u002FBenAAndrew\u002FVoice-Cloning-App\u002Flatest\u002Fartifacts\u002F0\u002Ftmp\u002Fbadges\u002Fcomment.svg?style=svg)](https:\u002F\u002Fcircleci.com\u002Fgh\u002FBenAAndrew\u002FVoice-Cloning-App)\n[![comment](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fcode%20style-black-000000.svg)](https:\u002F\u002Fgithub.com\u002Fpsf\u002Fblack)\n\nA Python\u002FPytorch app for easily synthesising human voices\n\n![Preview](preview.png \"Preview\")\n\n## [Documentation](https:\u002F\u002Fbenaandrew.github.io\u002FVoice-Cloning-App\u002F)\n\n## [Discord Server](https:\u002F\u002Fdiscord.gg\u002FwQd7zKCWxT)\n\n## [Video guide](https:\u002F\u002Fwww.youtube.com\u002Fplaylist?list=PLk5I7EvFL13GjBIDorh5yE1SaPGRG-i2l)\n\n## [Voice Sharing Hub](https:\u002F\u002Fvoice-sharing-hub.herokuapp.com\u002F)\n\n## [FAQ's](faqs.md)\n\n## System Requirements\n- **Windows 10 or Ubuntu 20.04+ operating system**\n- **5GB+ Disk space**\n- NVIDIA GPU with at least 4GB of memory & driver version 456.38+ (optional)\n\n## Key features\n- Automatic dataset generation (with support for subtitles and audiobooks)\n- Additional language support\n- Local & remote training\n- Easy train start\u002Fstop\n- Data importing\u002Fexporting\n- Multi GPU support\n\n## Manual Guides\n- [Installation](install.md)\n- [Building the dataset](dataset\u002Fdataset.md)\n- [Training](training\u002Ftraining.md)\n- [Synthesis](synthesis\u002Fsynthesis.md)\n- [Making changes](maintenance.md)\n\n## Future Improvements\n- Add support for Talknet\n- Add GTA alignment for Hifi-gan\n- Improved batch size estimation\n- AMD GPU support\n\n## Other resources\n- [Remote training notebook](https:\u002F\u002Fcolab.research.google.com\u002Fdrive\u002F1YbB_gA2_Rspmm1TDyEDueNXittmtWu1c?usp=sharing)\n- Try out existing voices at [uberduck.ai](https:\u002F\u002Fuberduck.ai\u002F) and [Vocodes](https:\u002F\u002Fvo.codes\u002F)\n- [Youtube data fetching](https:\u002F\u002Fcolab.research.google.com\u002Fdrive\u002F1_ulm1DKPOw8n0dHt8__2BR4d9WrWdWA4?usp=sharing) (created by Diskr33t#5880)\n- [Synthesize in Colab](https:\u002F\u002Fcolab.research.google.com\u002Fdrive\u002F18IJZZDW1NO7KOslg_WMOCrMeiqz9jOYF?usp=sharing) (created by mega b#6696)\n- [Generate youtube transcription](https:\u002F\u002Fcolab.research.google.com\u002Fdrive\u002F1KfAJig2jekpjJ5QS8Lpjy8sTd8w_ZuFv?usp=sharing) (created by mega b#6696)\n- [Wit.ai transcription](https:\u002F\u002Fcolab.research.google.com\u002Fdrive\u002F1i5hJRZVc0S-tgt5XM8kSoTu2nHBPOPrF#scrollTo=dk689PtThOjn)\n\n## Acknowledgements\nThis project uses a reworked version of [Tacotron2](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Ftacotron2). All rights for belong to NVIDIA and follow the requirements of their BSD-3 licence.\n\nAdditionally, the project uses [DSAlign](https:\u002F\u002Fgithub.com\u002Fmozilla\u002FDSAlign), [Silero](https:\u002F\u002Fgithub.com\u002Fsnakers4\u002Fsilero-models), [DeepSpeech](https:\u002F\u002Fgithub.com\u002Fmozilla\u002FDeepSpeech) & [hifi-gan](https:\u002F\u002Fgithub.com\u002Fjik876\u002Fhifi-gan).\n\nThank you to Dr. John Bustard at Queen's University Belfast for his support throughout the project.\n\nSupported by [uberduck.ai](https:\u002F\u002Fuberduck.ai\u002F), reach out to them for live model hosting.\n\nAlso a big thanks to the members of the [VocalSynthesis subreddit](https:\u002F\u002Fwww.reddit.com\u002Fr\u002FVocalSynthesis\u002F) for their feedback.\n\nFinally thank you to everyone raising issues and contributing to the project.\n","# 语音克隆应用\n[![CircleCI](https:\u002F\u002Fcircleci.com\u002Fgh\u002FBenAAndrew\u002FVoice-Cloning-App.svg?style=svg)](https:\u002F\u002Fcircleci.com\u002Fgh\u002FBenAAndrew\u002FVoice-Cloning-App)\n[![Discord](https:\u002F\u002Fimg.shields.io\u002Fdiscord\u002F833666557954883614.svg?style=flat-square)](https:\u002F\u002Fdiscord.gg\u002FwQd7zKCWxT)\n[![codecov](https:\u002F\u002Fcodecov.io\u002Fgh\u002FBenAAndrew\u002FVoice-Cloning-App\u002Fbranch\u002Fmain\u002Fgraph\u002Fbadge.svg?token=WC0LLZO3Z5)](https:\u002F\u002Fcodecov.io\u002Fgh\u002FBenAAndrew\u002FVoice-Cloning-App)\n[![comment](https:\u002F\u002Fcircleci.com\u002Fapi\u002Fv1.1\u002Fproject\u002Fgithub\u002FBenAAndrew\u002FVoice-Cloning-App\u002Flatest\u002Fartifacts\u002F0\u002Ftmp\u002Fbadges\u002Fcomment.svg?style=svg)](https:\u002F\u002Fcircleci.com\u002Fgh\u002FBenAAndrew\u002FVoice-Cloning-App)\n[![comment](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fcode%20style-black-000000.svg)](https:\u002F\u002Fgithub.com\u002Fpsf\u002Fblack)\n\n一个用于轻松合成人类声音的 Python\u002FPyTorch (深度学习框架) 应用程序\n\n![Preview](preview.png \"预览\")\n\n## [文档](https:\u002F\u002Fbenaandrew.github.io\u002FVoice-Cloning-App\u002F)\n\n## [Discord 服务器](https:\u002F\u002Fdiscord.gg\u002FwQd7zKCWxT)\n\n## [视频指南](https:\u002F\u002Fwww.youtube.com\u002Fplaylist?list=PLk5I7EvFL13GjBIDorh5yE1SaPGRG-i2l)\n\n## [语音共享中心](https:\u002F\u002Fvoice-sharing-hub.herokuapp.com\u002F)\n\n## [常见问题](faqs.md)\n\n## 系统要求\n- **Windows 10 或 Ubuntu 20.04+ 操作系统**\n- **5GB+ 磁盘空间**\n- NVIDIA GPU (英伟达显卡)，显存至少 4GB 且驱动版本 456.38+（可选）\n\n## 主要功能\n- 自动数据集生成（支持字幕和有声书）\n- 额外的语言支持\n- 本地与远程训练\n- 轻松开始\u002F停止训练\n- 数据导入\u002F导出\n- 多 GPU 支持\n\n## 手动指南\n- [安装](install.md)\n- [构建数据集](dataset\u002Fdataset.md)\n- [训练](training\u002Ftraining.md)\n- [合成](synthesis\u002Fsynthesis.md)\n- [进行修改](maintenance.md)\n\n## 未来改进\n- 添加对 Talknet 的支持\n- 为 Hifi-gan (高保真生成对抗网络) 添加 GTA (梯度时间对齐)\n- 改进批次大小估计\n- AMD GPU 支持\n\n## 其他资源\n- [远程训练笔记本](https:\u002F\u002Fcolab.research.google.com\u002Fdrive\u002F1YbB_gA2_Rspmm1TDyEDueNXittmtWu1c?usp=sharing)\n- 在 [uberduck.ai](https:\u002F\u002Fuberduck.ai\u002F) (语音平台) 和 [Vocodes](https:\u002F\u002Fvo.codes\u002F) (语音合成平台) 试用现有声音\n- [YouTube 数据获取](https:\u002F\u002Fcolab.research.google.com\u002Fdrive\u002F1_ulm1DKPOw8n0dHt8__2BR4d9WrWdWA4?usp=sharing)（由 Diskr33t#5880 创建）\n- [在 Colab (云端开发环境) 中合成](https:\u002F\u002Fcolab.research.google.com\u002Fdrive\u002F18IJZZDW1NO7KOslg_WMOCrMeiqz9jOYF?usp=sharing)（由 mega b#6696 创建）\n- [生成 YouTube 转录文本](https:\u002F\u002Fcolab.research.google.com\u002Fdrive\u002F1KfAJig2jekpjJ5QS8Lpjy8sTd8w_ZuFv?usp=sharing)（由 mega b#6696 创建）\n- [Wit.ai (语音识别服务) 转录](https:\u002F\u002Fcolab.research.google.com\u002Fdrive\u002F1i5hJRZVc0S-tgt5XM8kSoTu2nHBPOPrF#scrollTo=dk689PtThOjn)\n\n## 致谢\n本项目使用了经过重写的 [Tacotron2 (语音合成模型)](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Ftacotron2) 版本。所有权利均归属于 NVIDIA，并遵循其 BSD-3 许可证的要求。\n\n此外，本项目还使用了 [DSAlign](https:\u002F\u002Fgithub.com\u002Fmozilla\u002FDSAlign)、[Silero](https:\u002F\u002Fgithub.com\u002Fsnakers4\u002Fsilero-models)、[DeepSpeech](https:\u002F\u002Fgithub.com\u002Fmozilla\u002FDeepSpeech) 和 [hifi-gan](https:\u002F\u002Fgithub.com\u002Fjik876\u002Fhifi-gan)。\n\n感谢贝尔法斯特女王大学的 John Bustard 博士在整个项目期间提供的支持。\n\n由 [uberduck.ai](https:\u002F\u002Fuberduck.ai\u002F) 支持，如需实时模型托管请联系他们。\n\n同样非常感谢 [VocalSynthesis Reddit 板块](https:\u002F\u002Fwww.reddit.com\u002Fr\u002FVocalSynthesis\u002F) 的成员们提供的反馈。\n\n最后感谢所有提出问题并为项目做出贡献的人。","# Voice-Cloning-App 快速上手指南\n\n**Voice-Cloning-App** 是一个基于 Python 和 PyTorch 的开源应用，旨在轻松合成人类语音。支持数据集自动生成、多语言支持及本地\u002F远程训练。\n\n## 1. 环境准备\n\n在开始之前，请确保您的开发环境满足以下系统要求：\n\n- **操作系统**: Windows 10 或 Ubuntu 20.04+\n- **磁盘空间**: 至少 5GB\n- **GPU (可选但推荐)**: NVIDIA 显卡，显存至少 4GB，驱动版本 456.38+\n- **依赖环境**: Python, PyTorch\n\n> 💡 **国内开发者提示**：安装 Python 包时，建议使用国内镜像源以加速下载（例如清华源）：\n> ```bash\n> pip install -i https:\u002F\u002Fpypi.tuna.tsinghua.edu.cn\u002Fsimple \u003Cpackage_name>\n> ```\n\n## 2. 安装步骤\n\n### 克隆项目\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002FBenAAndrew\u002FVoice-Cloning-App.git\ncd Voice-Cloning-App\n```\n\n### 安装依赖\n本项目依赖较多，建议参考官方安装文档获取完整依赖列表。通常流程如下：\n\n```bash\npip install -r requirements.txt\n```\n\n> ⚠️ **注意**：具体的依赖配置和 CUDA 设置，请务必查阅项目中的 **[Installation](https:\u002F\u002Fbenaandrew.github.io\u002FVoice-Cloning-App\u002Finstall.html)** 文档。\n\n## 3. 基本使用\n\n本工具的核心工作流分为三个主要阶段：**构建数据集** -> **训练模型** -> **语音合成**。\n\n### 核心流程概览\n1.  **构建数据集 (Building the dataset)**\n    自动处理字幕或有声书，生成训练所需的数据集。\n    > 详细指南：[Building the dataset](dataset\u002Fdataset.md)\n\n2.  **训练模型 (Training)**\n    支持本地训练或多 GPU 训练，可控制训练的启动与停止。\n    > 详细指南：[Training](training\u002Ftraining.md)\n\n3.  **语音合成 (Synthesis)**\n    使用训练好的模型生成目标语音。\n    > 详细指南：[Synthesis](synthesis\u002Fsynthesis.md)\n\n### 替代方案：云端运行\n如果您没有本地 GPU 资源，可以使用 Google Colab 进行远程训练和合成：\n- [Remote training notebook](https:\u002F\u002Fcolab.research.google.com\u002Fdrive\u002F1YbB_gA2_Rspmm1TDyEDueNXittmtWu1c?usp=sharing)\n- [Synthesize in Colab](https:\u002F\u002Fcolab.research.google.com\u002Fdrive\u002F18IJZZDW1NO7KOslg_WMOCrMeiqz9jOYF?usp=sharing)\n\n## 4. 更多资源\n\n- **官方文档**: [Documentation](https:\u002F\u002Fbenaandrew.github.io\u002FVoice-Cloning-App\u002F)\n- **Discord 社区**: [Discord Server](https:\u002F\u002Fdiscord.gg\u002FwQd7zKCWxT)\n- **视频教程**: [Video guide](https:\u002F\u002Fwww.youtube.com\u002Fplaylist?list=PLk5I7EvFL13GjBIDorh5yE1SaPGRG-i2l)\n- **声音分享中心**: [Voice Sharing Hub](https:\u002F\u002Fvoice-sharing-hub.herokuapp.com\u002F)","独立游戏开发者小李正在制作一款多语言 RPG，急需为数十个 NPC 生成符合人设的对话语音，但受限于预算无法聘请专业配音团队。\n\n### 没有 Voice-Cloning-App 时\n- 寻找并录制真人配音成本高昂，单个角色的定制费用远超项目预算\n- 修改剧本后需重新录音，导致开发周期被严重拉长，沟通成本巨大\n- 自行搭建深度学习语音合成模型技术门槛高，环境配置复杂且易出错\n- 依赖在线 API 服务存在数据隐私风险，且长文本生成常出现延迟或中断\n\n### 使用 Voice-Cloning-App 后\n- 利用少量样本即可快速克隆特定音色，大幅降低配音成本，支持批量生成\n- 文本转语音流程自动化，修改台词无需重新录制，迭代效率提升显著\n- 支持本地训练与自动数据集生成，无需深厚算法背景即可上手，兼容多 GPU 加速\n- 数据完全在本地处理，保障了游戏资产的安全性与离线可用性，避免网络波动\n\nVoice-Cloning-App 通过简化语音合成流程，让非专业人士也能低成本实现高质量的角色配音定制。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fvoice-cloning-app_Voice-Cloning-App_4787a31b.png","voice-cloning-app","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002Fvoice-cloning-app_aaf3759e.png",null,"https:\u002F\u002Fgithub.com\u002Fvoice-cloning-app",[81,85,89,93,97,101],{"name":82,"color":83,"percentage":84},"Python","#3572A5",80.7,{"name":86,"color":87,"percentage":88},"HTML","#e34c26",11.3,{"name":90,"color":91,"percentage":92},"JavaScript","#f1e05a",6.6,{"name":94,"color":95,"percentage":96},"CSS","#663399",1,{"name":98,"color":99,"percentage":100},"Shell","#89e051",0.2,{"name":102,"color":103,"percentage":104},"Dockerfile","#384d54",0.1,1442,239,"2026-03-29T13:38:52","BSD-3-Clause","Windows, Linux","需要 NVIDIA GPU，显存 4GB+，驱动版本 456.38+（可选）","未说明",{"notes":113,"python":111,"dependencies":114},"需至少 5GB 磁盘空间；GPU 为可选配置；支持多 GPU 训练及远程训练；项目基于 NVIDIA Tacotron2 重构。",[115,116,117,118,119,120],"torch","tacotron2","dsalign","silero-models","deepspeech","hifi-gan",[55,13],[123,124,125,126,127,128,116],"python","tts","text-to-speech","pytorch","deep-learning","voice-cloning",4,"2026-03-27T02:49:30.150509","2026-04-06T07:14:48.655135",[133,138,142,146,151,156],{"id":134,"question_zh":135,"answer_zh":136,"source_url":137},2118,"浏览器窗口自动打开后卡住或无响应怎么办？","尝试手动关闭自动弹出的浏览器窗口，并在 Chrome 等浏览器中手动访问 localhost:5000。有用户反馈直接让浏览器窗口挂着会导致问题，手动选择浏览器或默认使用 Chrome 可解决。维护者指出处理非默认浏览器较复杂，但确认应用本身不需要互联网连接即可运行。","https:\u002F\u002Fgithub.com\u002Fvoice-cloning-app\u002FVoice-Cloning-App\u002Fissues\u002F11",{"id":139,"question_zh":140,"answer_zh":141,"source_url":137},2119,"训练过程中检查点保存频率太低，如何防止数据丢失？","建议在 train.py 中降低 ITERS_PER_CHECKPOINT 的设置值（当前默认为 1000）。有用户在 GTX 1080 上反馈默认设置下约每小时才保存一次，容易因意外中断丢失进度。建议增加手动保存功能或提高检查点频率以便更频繁地保存实验状态。",{"id":143,"question_zh":144,"answer_zh":145,"source_url":137},2120,"应用是否需要联网才能运行？","不需要。维护者明确指出，尽管应用通过浏览器界面运行，但它不需要互联网连接即可正常工作。因此，本地局域网内的部署也是可行的。",{"id":147,"question_zh":148,"answer_zh":149,"source_url":150},2121,"训练时出现 GPU 显存溢出（Memory exhausted）如何解决？","首先检查是否有大量后台进程占用了显存，关闭不必要的程序。其次，确保使用最新版本的应用（如 0.4 版本后已修复相关问题）。如果仍存在问题，可尝试减少 Epoch 数量或迭代次数，但主要解决方案是释放系统资源。","https:\u002F\u002Fgithub.com\u002Fvoice-cloning-app\u002FVoice-Cloning-App\u002Fissues\u002F7",{"id":152,"question_zh":153,"answer_zh":154,"source_url":155},2122,"训练任务突然崩溃或进度停滞不前是什么原因？","这通常是由于线程处理机制导致的。该问题已在 0.6 版本中修复。维护者计划将所有进程切换为子进程（subprocesses）以避免此类问题。如果遇到此情况，建议升级应用至最新版本。","https:\u002F\u002Fgithub.com\u002Fvoice-cloning-app\u002FVoice-Cloning-App\u002Fissues\u002F18",{"id":157,"question_zh":158,"answer_zh":159,"source_url":160},2123,"上传文件时出现 HTTP 503 错误或无法加载 Silero 模型怎么办？","这通常是因为外部依赖服务（Silero 模型托管）暂时不可用导致的网络错误，并非应用本身故障。建议稍后重试。如果持续失败，可检查网络连接或关注相关模型库的 GitHub 仓库。有用户反馈重新安装 Python 无效，需等待服务恢复。","https:\u002F\u002Fgithub.com\u002Fvoice-cloning-app\u002FVoice-Cloning-App\u002Fissues\u002F48",[162,167,172,177,182,187,192,197,202,207,212,217,222,227,232,237,242,247,252,257],{"id":163,"version":164,"summary_zh":165,"released_at":166},101585,"v1.1.1","Build with GPU support: https:\u002F\u002Fmega.nz\u002Ffile\u002F4hpgFBAb#6GP3p0n-s5v9KIXmFRqxDst7BPaNhEy14JMlEe0aopY\r\n\r\nNotes:\r\nImprove transcription error logging\r\nUpdate dataset info when manually labelling a clip\r\nFix quotation problem in synthesis text","2022-02-07T15:33:10",{"id":168,"version":169,"summary_zh":170,"released_at":171},101586,"v1.1.0","Build with GPU support: https:\u002F\u002Fmega.nz\u002Ffile\u002F0woADQSS#vud7UO7Pi-wNsa6eawqqdzcqNoCNGzH50XHDlbQnd0E\r\n\r\nNotes:\r\nAdd custom vocoder training\r\nFix data path issue to ensure all files are in the data folder","2021-12-09T20:09:32",{"id":173,"version":174,"summary_zh":175,"released_at":176},101587,"v1.0.4","Build with GPU support: https:\u002F\u002Fmega.nz\u002Ffile\u002Fh8wQBDJY#j4jq3PVj4LDhMPeJJwwhod1bVuve9Sf3G6NyYzo-mZg\r\n\r\nNotes:\r\nFix unlabelled clip playback\r\nAdd new remote training notebook\r\nImprove invalid symbols error message (thanks to @SirBitesalot)\r\nFix symbols selection for non-English dataset creation (thanks to @SirBitesalot)","2021-11-29T18:29:36",{"id":178,"version":179,"summary_zh":180,"released_at":181},101588,"v1.0.3","Build with GPU support: https:\u002F\u002Fmega.nz\u002Ffile\u002FQ5Ak2BZL#-AQTJyHO-wh0sGC5IcCX6Hjce6WQvhT7OXaueSZ2ztg\r\n\r\nNotes:\r\nFixed issue with invalid clips being added for manual labelling\r\nFixed multi-line synthesis results folder naming (thanks @Marclass)\r\nAdded support for custom languages to text cleaning to improve non-English dataset quality\r\nImproved dataset validation and errors\r\n\r\n","2021-11-24T22:57:23",{"id":183,"version":184,"summary_zh":185,"released_at":186},101589,"v1.0.2","Build with GPU support: https:\u002F\u002Fmega.nz\u002Ffile\u002FFkgGASaL#q8hn70t6zn6m_a9UwwwI2RWbwz3BqzaJ6b7nfxucUbg\r\n\r\nNotes:\r\nRemove pre-trained model buildup\r\nAdd manage dataset option to view dataset info & label unlabelled clips","2021-10-16T21:28:38",{"id":188,"version":189,"summary_zh":190,"released_at":191},101590,"v1.0.1","Build with GPU support no longer available\r\n\r\nNotes:\r\nFix training from checkpoint issue\r\nFix alignment GIF bug\r\nUpdate navigation bar (including links to docs & discord)","2021-10-05T22:14:14",{"id":193,"version":194,"summary_zh":195,"released_at":196},101591,"v1.0.0","Build with GPU support: https:\u002F\u002Fmega.nz\u002Ffile\u002FxlgXlQSD#Ls1BN_CDDyoL5cqjzBhrh63RTQAa6fUv5DYnDY0Jowo\r\n\r\nNotes:\r\nGenerate an alignment graph in training to visualise how the model is performing against a test sentence\r\nExport timelapse of alignments to visualise improvement over time\r\nAdd sentence length recommendation to synthesis","2021-09-18T19:28:29",{"id":198,"version":199,"summary_zh":200,"released_at":201},101592,"v0.9.9","Build with GPU support no longer available\r\n\r\nNotes:\r\nPackage FFmpeg & Silero with the executable to avoid additional downloads","2021-09-15T15:29:44",{"id":203,"version":204,"summary_zh":205,"released_at":206},101593,"v0.9.8","Build with GPU support no longer available\r\n\r\nNotes:\r\nImprove symbol weights transfer to improve training from an existing model (thanks to @CookiePPP)\r\nAdd attention scoring to training\r\nFix dataset info generation when extending existing datasets\r\nFix issue with checkpoint selection on the training page","2021-09-08T15:32:21",{"id":208,"version":209,"summary_zh":210,"released_at":211},101594,"v0.9.7","Build with GPU support no longer available\r\n\r\nNotes:\r\nAdd paragraph synthesis option with automatic sentence splitting\r\nRemove clip length requirement in dataset importing","2021-08-31T17:03:50",{"id":213,"version":214,"summary_zh":215,"released_at":216},101595,"v0.9.6","Build with GPU support no longer available\r\n\r\nNotes:\r\nAdd clip combiner and duration options to dataset building\r\nFix blank line handling in synthesis\r\nFix checkpoint exporting\r\nStandardise alignment JSON","2021-08-29T12:23:10",{"id":218,"version":219,"summary_zh":220,"released_at":221},101596,"v0.9.5","Build with GPU support no longer available\r\n\r\nNotes:\r\nRemove waveglow\r\nUpdate to CUDA 11.1 to support 30 series GPU's","2021-08-18T16:01:14",{"id":223,"version":224,"summary_zh":225,"released_at":226},101597,"v0.9.4","Add checkpoint backup system\r\nAdd subtitle support","2021-08-10T10:22:46",{"id":228,"version":229,"summary_zh":230,"released_at":231},101598,"v0.9.3","Add train\u002Ftest split slider to training\r\nNew multi-line synthesis feature\r\nAdd max_decoder_steps slider to synthesis","2021-07-28T15:23:29",{"id":233,"version":234,"summary_zh":235,"released_at":236},101599,"v0.9.2","Remove synonym suggestion\r\nMove vocoder upload to settings\r\nFix checkpoint selection in training","2021-07-27T16:42:42",{"id":238,"version":239,"summary_zh":240,"released_at":241},101600,"v0.9.1","Fix English synthesis\r\nEnable checkpoint selection for export, training & synthesis","2021-07-25T20:31:40",{"id":243,"version":244,"summary_zh":245,"released_at":246},101601,"v0.9","Added support for other languages","2021-07-24T21:10:19",{"id":248,"version":249,"summary_zh":250,"released_at":251},101602,"v0.8.4","Update Kindle extraction guide","2021-07-23T13:15:31",{"id":253,"version":254,"summary_zh":255,"released_at":256},101603,"v0.8.3","Add advanced training options for enabling\u002Fdisabling checkpoint overwriting & multi GPU","2021-07-17T18:24:33",{"id":258,"version":259,"summary_zh":260,"released_at":261},101604,"v0.8.2","Update Pytorch version and revert CUDA to 10.2","2021-07-02T20:24:46"]