[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"site-home-live-entry-setting":3,"site-header-contests":8,"news-detail-news-47":49,"site-header-camps":94},{"enabled":4,"title":5,"statusText":5,"buttonText":5,"campSlug":5,"stage":6,"classroomId":7,"hasClassroom":4},false,"",0,null,{"total":9,"list":10,"pageNum":44,"pageSize":45,"size":9,"startRow":44,"endRow":9,"pages":44,"prePage":6,"nextPage":6,"isFirstPage":46,"isLastPage":46,"hasPreviousPage":4,"hasNextPage":4,"navigatePages":47,"navigatepageNums":48,"navigateFirstPage":44,"navigateLastPage":44},3,[11,25,35],{"slug":12,"title":13,"logo":7,"liveUrl":14,"prize":15,"sponsor":16,"description":17,"organizer":18,"startTime":19,"endTime":20,"registrationDeadline":21,"status":22,"dashboardVisibility":23,"createdAt":24},"spring2026","2026春季启元人工智能大赛","https:\u002F\u002Fwww.bilibili.com\u002Fblackboard\u002Flive\u002Flive-activity-player.html?cid=1822703816&quality=0","200,000","麒麟软件有限公司","“2026春季启元人工智能大赛”是由启元实验室主办的面向全国人工智能领域开发者、科研人员及高校学生的权威赛事。大赛旨在促进与培养具备工程实践能力和创新能力的优秀人工智能领域开发者、推广国产软硬件技术与生态、助力国家人工智能产业高质量发展。\n各参赛选手需在规定的比赛周期内，依据赛事规则选定赛题并完成相应挑战，优胜者将共同角逐总额20万元的赛事奖金池，并且还将有机会获得实习和知名企业优质岗位内推资格。","启元实验室","2026-04-19 16:00:00","2026-07-13 16:00:00","2026-06-16 16:00:00","ongoing","private","2026-04-16 08:00:00",{"slug":26,"title":27,"logo":7,"liveUrl":14,"prize":28,"sponsor":7,"description":29,"organizer":18,"startTime":30,"endTime":31,"registrationDeadline":32,"status":33,"dashboardVisibility":23,"createdAt":34},"autumn2025","2025秋季启元人工智能大赛","150,000","“2025秋季启元人工智能大赛”是由启元实验室主办的面向全国人工智能领域开发者、科研人员及高校学生的权威赛事。大赛旨在促进与培养具备工程实践能力和创新能力的优秀人工智能领域开发者、推广国产软硬件技术与生态、助力国家人工智能产业高质量发展。","2025-11-08 08:00:00","2026-01-13 08:00:00","2025-12-30 08:00:00","ended","2025-11-03 08:00:00",{"slug":36,"title":37,"logo":7,"liveUrl":14,"prize":38,"sponsor":7,"description":39,"organizer":18,"startTime":40,"endTime":41,"registrationDeadline":42,"status":33,"dashboardVisibility":23,"createdAt":43},"summer2025","启元人工智能大赛(系统方向)","80,000","启元人工智能大赛由启元实验室主办，其中系统赛道基于启元实验室九源软件栈的相关开源项目，通过拉动开源社区力量，一方面促进国产智能计算生态的发展，另一方面也为人工智能系统方向的学生及从业者提供一个实践平台。","2025-07-08 08:00:00","2025-08-25 08:00:00","2025-07-22 08:00:00","2025-06-22 08:00:00",1,5,true,8,[44],{"newsDetail":50,"processedData":71},{"publishTime":51,"img":52,"keywords":53,"treeType":44,"showStatus":44,"updateTime":54,"type":44,"title":55,"content":56,"nextArticle":57,"createTime":54,"sortOrder":62,"canonicalPath":63,"id":64,"contentType":65,"prevArticle":66},"2026-06-10 08:00:00","https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002Fportal_news\u002Fnews_logo_fm.jpg?Expires=1792752856&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=nlIf34vJUNL8fONDIdbvxhU5u%2Fc%3D","InfiniLM,PD 分离,异构部署,RDMA 通信,大模型推理,国产加速卡","2026-10-22 20:38:11","九源大模型推理框架 InfiniLM 在国产异构 PD 分离场景取得阶段性进展",">近日，InfiniTensor 团队基于**九源大模型推理框架 InfiniLM**，与 Mooncake 及中科院计算所团队协同开展相关工作，在国产异构 PD 分离推理场景中取得阶段性进展。\r\n\r\n\r\n随着大模型推理规模持续增长，Prefill-Decode（PD）分离架构正逐渐成为提升系统吞吐、优化资源利用的重要技术方向。近期，InfiniTensor 团队基于 InfiniLM，在国产异构 PD 分离能力建设方面取得了新的进展。\r\n\r\n目前，InfiniLM 已支持英伟达、沐曦、摩尔线程、天数智芯等不同计算平台的混合部署，可根据业务需求灵活配置 Prefill 节点与 Decode 节点，构建由多种加速卡协同参与的异构 PD 分离推理集群。系统支持 **“x 种加速卡 m 个 P 节点 + y 种加速卡 n 个 D 节点”** 的部署模式，并通过 RDMA 通信实现高效的跨节点协作。\r\n\r\n目前，相关功能已完成核心验证，正在持续推进性能优化、稳定性验证及工程化整理工作。\r\n\r\n![](https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002Fportal_news\u002Fnews_47_zw01.jpg?Expires=1792752856&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=%2BP4Rq292axxj8TAgnpK4ULF6jxs%3D)\r\n\r\n\r\n## 打破算力壁垒，让每一份算力物尽其用\r\n\r\n随着大模型推理规模不断扩大，Prefill 与 Decode 两个阶段在计算特征上的差异逐渐受到关注。Prefill 阶段以计算密集型负载为主，而 Decode 阶段则更依赖显存访问与数据传输。将两类负载部署在统一资源池中，往往难以充分发挥不同硬件的特点。\r\n\r\n针对这一问题，InfiniLM 对 PD（Prefill-Decode）分离架构进行了探索与实践，通过任务解耦和高速互联机制，使 Prefill 与 Decode 节点能够独立部署、协同工作：\r\n\r\n- **Prefill 节点**：负责输入 Prompt 的批量计算与 KV Cache 生成；\r\n- **Decode 节点**：基于已生成的 KV Cache 执行逐 Token 推理；\r\n- **跨节点协同**：通过 RDMA 网络传输 KV Cache，降低跨节点数据交换开销，提升异构节点间的协同效率。\r\n\r\n## 核心技术亮点一览\r\n\r\n上述进展的背后，是 InfiniLM 在系统架构、调度机制与通信框架等方面的深入探索：\r\n\r\n### PD 感知的 KV Connector 分层架构：\r\n参考业界先进架构的Scheduler\u002FWorker 分离设计，实现调度逻辑与数据传输的完美解耦。目前已经支持 Mooncake高速传输后端，并预留了接入其他通信后端的接口。\r\n\r\n\r\n![](https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002Fportal_news\u002Fnews_47_zw02.jpg?Expires=1792752856&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=%2BqjguJ%2FhYXUyjOCJziRbdUd9DWU%3D)\r\n\r\n\r\n### 基于 Mooncake 的 RDMA 传输\r\n使用 Mooncake TransferEngine 实现 GPUDirect RDMA，KV Cache 在 P\u002FD 节点 GPU 显存间直传，将传输延迟大幅度降低。\r\n\r\n### 智能两阶段调度与块管理\r\n引入 WAITING_FOR_REMOTE_KVS 新状态，allocate_slots等函数精确协同 处理 PD 块分配与 Prefix Cache，缓存命中词元可享“零传输”加速。\r\n\r\n\r\n![](https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002Fportal_news\u002Fnews_47_zw03.jpg?Expires=1792752856&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=ZCq%2B%2FrBxNrm%2F%2FR%2Fb39OMMQpzNLI%3D)\r\n\r\n\r\n### 高可用异常处理机制\r\n全链路覆盖传输失败、节点异常等场景。实现请求取消的级联通知、KV 传输失败后自动回退重算等机制，确保系统鲁棒性。\r\n\r\n\r\n## 从能力验证到持续演进\r\n此次国产异构 PD 分离能力的实现，是 InfiniTensor、Mooncake 与中科院计算所等团队协同推进的阶段性成果。在约两个月的时间里，团队围绕架构设计、通信机制、调度策略与工程实现等方面持续迭代，完成了相关能力的开发与验证。\r\n\r\n随着局部优化和代码整理工作的深入，InfiniLM 即将完成多机 RDMA 端到端联调，并推出稳定版本。未来，我们将继续朝着**动态扩缩容、跨集群 KV Cache 共享**等方向演进，为企业级大模型推理部署提供更强大、更灵活、更高效的基础设施。\r\n\r\n我们也期待与更多生态伙伴共同推进国产 AI 推理技术的发展，持续探索异构算力协同与大模型推理优化的更多可能性。\r\n\r\n## 关注与交流\r\n\r\n![](https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002Fportal_news\u002Fnews_ewm02.jpg?Expires=1792752856&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=PWOcTn3Z9OJvJpDh%2BsqT%2B6vsPsQ%3D)",{"publishTime":58,"createTime":54,"canonicalPath":59,"id":60,"title":61},"2026-06-02 08:00:00","\u002Fnews\u002Fnews-46","news-46","InfiniCCL v0.1.0 发布：一个面向异构集群的统一通信库",27,"\u002Fnews\u002Fnews-47","news-47","markdown",{"publishTime":67,"createTime":54,"canonicalPath":68,"id":69,"title":70},"2026-06-17 08:00:00","\u002Fnews\u002Fnews-48","news-48","九源智能计算生态联合体走进兰州大学，开启训练营与人工智能大赛招募",{"htmlContent":72,"headings":73},"\u003Cblockquote>\n\u003Cp>近日，InfiniTensor 团队基于\u003Cstrong>九源大模型推理框架 InfiniLM\u003C\u002Fstrong>，与 Mooncake 及中科院计算所团队协同开展相关工作，在国产异构 PD 分离推理场景中取得阶段性进展。\u003C\u002Fp>\n\u003C\u002Fblockquote>\n\u003Cp>随着大模型推理规模持续增长，Prefill-Decode（PD）分离架构正逐渐成为提升系统吞吐、优化资源利用的重要技术方向。近期，InfiniTensor 团队基于 InfiniLM，在国产异构 PD 分离能力建设方面取得了新的进展。\u003C\u002Fp>\n\u003Cp>目前，InfiniLM 已支持英伟达、沐曦、摩尔线程、天数智芯等不同计算平台的混合部署，可根据业务需求灵活配置 Prefill 节点与 Decode 节点，构建由多种加速卡协同参与的异构 PD 分离推理集群。系统支持 \u003Cstrong>“x 种加速卡 m 个 P 节点 + y 种加速卡 n 个 D 节点”\u003C\u002Fstrong> 的部署模式，并通过 RDMA 通信实现高效的跨节点协作。\u003C\u002Fp>\n\u003Cp>目前，相关功能已完成核心验证，正在持续推进性能优化、稳定性验证及工程化整理工作。\u003C\u002Fp>\n\u003Cp>\u003Cimg src=\"https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002Fportal_news\u002Fnews_47_zw01.jpg?Expires=1792752856&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=%2BP4Rq292axxj8TAgnpK4ULF6jxs%3D\" alt=\"\">\u003C\u002Fp>\n\u003Ch2 id=\"打破算力壁垒-让每一份算力物尽其用\">\u003Ca class=\"anchor\" href=\"#打破算力壁垒-让每一份算力物尽其用\" aria-label=\"跳转到标题：打破算力壁垒，让每一份算力物尽其用\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>打破算力壁垒，让每一份算力物尽其用\u003C\u002Fh2>\u003Cp>随着大模型推理规模不断扩大，Prefill 与 Decode 两个阶段在计算特征上的差异逐渐受到关注。Prefill 阶段以计算密集型负载为主，而 Decode 阶段则更依赖显存访问与数据传输。将两类负载部署在统一资源池中，往往难以充分发挥不同硬件的特点。\u003C\u002Fp>\n\u003Cp>针对这一问题，InfiniLM 对 PD（Prefill-Decode）分离架构进行了探索与实践，通过任务解耦和高速互联机制，使 Prefill 与 Decode 节点能够独立部署、协同工作：\u003C\u002Fp>\n\u003Cul>\n\u003Cli>\u003Cstrong>Prefill 节点\u003C\u002Fstrong>：负责输入 Prompt 的批量计算与 KV Cache 生成；\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Decode 节点\u003C\u002Fstrong>：基于已生成的 KV Cache 执行逐 Token 推理；\u003C\u002Fli>\n\u003Cli>\u003Cstrong>跨节点协同\u003C\u002Fstrong>：通过 RDMA 网络传输 KV Cache，降低跨节点数据交换开销，提升异构节点间的协同效率。\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch2 id=\"核心技术亮点一览\">\u003Ca class=\"anchor\" href=\"#核心技术亮点一览\" aria-label=\"跳转到标题：核心技术亮点一览\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>核心技术亮点一览\u003C\u002Fh2>\u003Cp>上述进展的背后，是 InfiniLM 在系统架构、调度机制与通信框架等方面的深入探索：\u003C\u002Fp>\n\u003Ch3 id=\"pd-感知的-kv-connector-分层架构\">\u003Ca class=\"anchor\" href=\"#pd-感知的-kv-connector-分层架构\" aria-label=\"跳转到标题：PD 感知的 KV Connector 分层架构：\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>PD 感知的 KV Connector 分层架构：\u003C\u002Fh3>\u003Cp>参考业界先进架构的Scheduler\u002FWorker 分离设计，实现调度逻辑与数据传输的完美解耦。目前已经支持 Mooncake高速传输后端，并预留了接入其他通信后端的接口。\u003C\u002Fp>\n\u003Cp>\u003Cimg src=\"https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002Fportal_news\u002Fnews_47_zw02.jpg?Expires=1792752856&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=%2BqjguJ%2FhYXUyjOCJziRbdUd9DWU%3D\" alt=\"\">\u003C\u002Fp>\n\u003Ch3 id=\"基于-mooncake-的-rdma-传输\">\u003Ca class=\"anchor\" href=\"#基于-mooncake-的-rdma-传输\" aria-label=\"跳转到标题：基于 Mooncake 的 RDMA 传输\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>基于 Mooncake 的 RDMA 传输\u003C\u002Fh3>\u003Cp>使用 Mooncake TransferEngine 实现 GPUDirect RDMA，KV Cache 在 P\u002FD 节点 GPU 显存间直传，将传输延迟大幅度降低。\u003C\u002Fp>\n\u003Ch3 id=\"智能两阶段调度与块管理\">\u003Ca class=\"anchor\" href=\"#智能两阶段调度与块管理\" aria-label=\"跳转到标题：智能两阶段调度与块管理\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>智能两阶段调度与块管理\u003C\u002Fh3>\u003Cp>引入 WAITING_FOR_REMOTE_KVS 新状态，allocate_slots等函数精确协同 处理 PD 块分配与 Prefix Cache，缓存命中词元可享“零传输”加速。\u003C\u002Fp>\n\u003Cp>\u003Cimg src=\"https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002Fportal_news\u002Fnews_47_zw03.jpg?Expires=1792752856&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=ZCq%2B%2FrBxNrm%2F%2FR%2Fb39OMMQpzNLI%3D\" alt=\"\">\u003C\u002Fp>\n\u003Ch3 id=\"高可用异常处理机制\">\u003Ca class=\"anchor\" href=\"#高可用异常处理机制\" aria-label=\"跳转到标题：高可用异常处理机制\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>高可用异常处理机制\u003C\u002Fh3>\u003Cp>全链路覆盖传输失败、节点异常等场景。实现请求取消的级联通知、KV 传输失败后自动回退重算等机制，确保系统鲁棒性。\u003C\u002Fp>\n\u003Ch2 id=\"从能力验证到持续演进\">\u003Ca class=\"anchor\" href=\"#从能力验证到持续演进\" aria-label=\"跳转到标题：从能力验证到持续演进\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>从能力验证到持续演进\u003C\u002Fh2>\u003Cp>此次国产异构 PD 分离能力的实现，是 InfiniTensor、Mooncake 与中科院计算所等团队协同推进的阶段性成果。在约两个月的时间里，团队围绕架构设计、通信机制、调度策略与工程实现等方面持续迭代，完成了相关能力的开发与验证。\u003C\u002Fp>\n\u003Cp>随着局部优化和代码整理工作的深入，InfiniLM 即将完成多机 RDMA 端到端联调，并推出稳定版本。未来，我们将继续朝着\u003Cstrong>动态扩缩容、跨集群 KV Cache 共享\u003C\u002Fstrong>等方向演进，为企业级大模型推理部署提供更强大、更灵活、更高效的基础设施。\u003C\u002Fp>\n\u003Cp>我们也期待与更多生态伙伴共同推进国产 AI 推理技术的发展，持续探索异构算力协同与大模型推理优化的更多可能性。\u003C\u002Fp>\n\u003Ch2 id=\"关注与交流\">\u003Ca class=\"anchor\" href=\"#关注与交流\" aria-label=\"跳转到标题：关注与交流\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>关注与交流\u003C\u002Fh2>\u003Cp>\u003Cimg src=\"https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002Fportal_news\u002Fnews_ewm02.jpg?Expires=1792752856&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=PWOcTn3Z9OJvJpDh%2BsqT%2B6vsPsQ%3D\" alt=\"\">\u003C\u002Fp>\n",[74,78,80,83,86,88,90,92],{"id":75,"text":76,"level":77},"打破算力壁垒-让每一份算力物尽其用","打破算力壁垒，让每一份算力物尽其用",2,{"id":79,"text":79,"level":77},"核心技术亮点一览",{"id":81,"text":82,"level":9},"pd-感知的-kv-connector-分层架构","PD 感知的 KV Connector 分层架构：",{"id":84,"text":85,"level":9},"基于-mooncake-的-rdma-传输","基于 Mooncake 的 RDMA 传输",{"id":87,"text":87,"level":9},"智能两阶段调度与块管理",{"id":89,"text":89,"level":9},"高可用异常处理机制",{"id":91,"text":91,"level":77},"从能力验证到持续演进",{"id":93,"text":93,"level":77},"关注与交流",{"total":45,"list":95,"pageNum":44,"pageSize":45,"size":45,"startRow":44,"endRow":45,"pages":44,"prePage":6,"nextPage":6,"isFirstPage":46,"isLastPage":46,"hasPreviousPage":4,"hasNextPage":4,"navigatePages":47,"navigatepageNums":148,"navigateFirstPage":44,"navigateLastPage":44},[96,107,118,129,139],{"id":97,"slug":98,"title":99,"logo":100,"publicLiveEmbedUrl":14,"description":101,"introduction":102,"organizer":18,"coOrganizer":5,"sponsor":16,"startTime":103,"endTime":104,"status":22,"dashboardVisibility":105,"createdAt":103,"updatedAt":106},"019eeeff-6b00-7d55-a629-efafded3e232","summer2026","2026年夏季InfiniTensor大模型与人工智能系统训练营","https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fuploads\u002F2026-06\u002F019eef7f-7816-7828-9309-79c2c61dd568.jpg?Expires=1792752856&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=SqASSorKbGqbBwQFQ5RjfAe6ISQ%3D","由启元实验室发起的InfiniTensor大模型与人工智能系统训练营，面向开源社区、高校及产业界伙伴开放。活动以「普及AI全栈知识技术、培养工程实践与创新型人才、推广国产软硬件技术生态」为核心目标，助力国家人工智能产业高质量发展。","...","2026-06-22 08:00:00","2026-10-19 08:00:00","public","2026-10-12 08:00:00",{"id":108,"slug":109,"title":110,"logo":111,"publicLiveEmbedUrl":7,"description":112,"introduction":102,"organizer":18,"coOrganizer":5,"sponsor":113,"startTime":114,"endTime":115,"status":33,"dashboardVisibility":23,"createdAt":116,"updatedAt":117},"019af3f3-e06a-7245-aa8d-37dc99a69c73","winter2025","2025年冬季InfiniTensor大模型与人工智能系统训练营","https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fuploads\u002F2025-12\u002F019afc69-b7ca-7805-91fb-605429671c01.png?Expires=1792752856&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=R5JKr8bO3hHaCXaP1wzWmJLE%2F1s%3D","由启元实验室发起的InfiniTensor大模型与人工智能系统训练营，面向开源社区、高校及产业界伙伴开放。活动以「普及AI全栈知识技术、培养工程实践与创新型人才、推广国产软硬件技术生态」为核心目标，助力国家人工智能产业高质量发展。 \n","新华三技术有限公司","2025-12-25 08:00:00","2026-03-27 08:00:00","2025-12-06 08:00:00","2026-10-22 20:24:30",{"id":119,"slug":36,"title":120,"logo":121,"publicLiveEmbedUrl":7,"description":122,"introduction":102,"organizer":18,"coOrganizer":5,"sponsor":123,"startTime":124,"endTime":125,"status":126,"dashboardVisibility":23,"createdAt":127,"updatedAt":128},"01980342-208c-7803-b3f3-19e1d31960ba","2025年夏季InfiniTensor大模型与人工智能系统训练营","https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fuploads\u002F2025-07\u002F0198045e-28fe-78f0-95fe-23835ce58395.jpg?Expires=1792752856&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=4gC4JkvFTMfNf4MMYpG%2BY8amgGQ%3D","InfiniTensor大模型与人工智能系统训练营由启元实验室及清华大学 AI 系统方向团队发起，面向开源社区、高校和产业界，普及人工智能各方面知识和技术，培养相关人才。训练营联合开源项目开发者、学术界大牛、工业界资深工程师参与授课与项目指导。课程采用「线上直播授课 + 在线答疑 + 项目实践辅导 」的方式进行，确保每位同学都能获得充分成长。训练营全程免费，通过视频、直播课、习题和项目实践，在人工智能飞速发展的背景下，为学员提供一窥深层技术的机会。","北京积算科技有限公司","2025-07-14 08:00:00","2025-09-30 08:00:00","upcoming","2025-07-13 08:00:00","2026-10-22 20:24:58",{"id":130,"slug":131,"title":132,"logo":133,"publicLiveEmbedUrl":7,"description":134,"introduction":102,"organizer":18,"coOrganizer":5,"sponsor":5,"startTime":135,"endTime":136,"status":126,"dashboardVisibility":23,"createdAt":137,"updatedAt":138},"01980cf9-a7d8-7796-ae86-0bbe05334add","winter2024","2024年冬季InfiniTensor大模型与人工智能系统训练营","https:\u002F\u002Ftesting-cdn.infinitensor.com\u002Fuploads\u002F2025-07\u002F01981269-ce23-7624-a287-9f07dd0d85f6.jpg","InfiniTensor大模型与人工智能系统训练营由清华大学发起，面向开源社区、高校和产业界，普及人工智能各方面知识和技术，培养相关人才。训练营全程免费，通过视频、直播课、习题和项目实践，在人工智能飞速发展的背景下，为学员提供一窥深层技术的机会。报名学员请阅读“导学”课程页的训练营介绍。","2025-01-05 08:00:00","2025-03-22 08:00:00","2025-07-15 08:00:00","2026-10-22 20:24:39",{"id":140,"slug":141,"title":142,"logo":143,"publicLiveEmbedUrl":7,"description":144,"introduction":102,"organizer":18,"coOrganizer":5,"sponsor":5,"startTime":145,"endTime":146,"status":126,"dashboardVisibility":23,"createdAt":137,"updatedAt":147},"01980c20-363a-7dc3-8f5a-58a68e08dcc4","summer2024","2024年夏季InfiniTensor大模型与人工智能系统训练营","https:\u002F\u002Ftesting-cdn.infinitensor.com\u002Fuploads\u002F2025-07\u002F01981269-fbfa-788a-afcd-17fa9be74ff9.jpg","InfiniTensor大模型与人工智能系统训练营由清华大学发起，面向开源社区和高校，普及人工智能各方面知识和技术，培养相关人才。训练营全程免费，通过课程、习题和项目实践，在人工智能飞速发展的背景下，为学员提供一窥深层技术的机会。","2024-07-20 08:00:00","2024-09-29 08:00:00","2026-10-22 20:24:49",[44]]