[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-determined-ai--determined":3,"tool-determined-ai--determined":61},[4,18,26,36,44,53],{"id":5,"name":6,"github_repo":7,"description_zh":8,"stars":9,"difficulty_score":10,"last_commit_at":11,"category_tags":12,"status":17},4358,"openclaw","openclaw\u002Fopenclaw","OpenClaw 是一款专为个人打造的本地化 AI 助手，旨在让你在自己的设备上拥有完全可控的智能伙伴。它打破了传统 AI 助手局限于特定网页或应用的束缚，能够直接接入你日常使用的各类通讯渠道，包括微信、WhatsApp、Telegram、Discord、iMessage 等数十种平台。无论你在哪个聊天软件中发送消息，OpenClaw 都能即时响应，甚至支持在 macOS、iOS 和 Android 设备上进行语音交互，并提供实时的画布渲染功能供你操控。\n\n这款工具主要解决了用户对数据隐私、响应速度以及“始终在线”体验的需求。通过将 AI 部署在本地，用户无需依赖云端服务即可享受快速、私密的智能辅助，真正实现了“你的数据，你做主”。其独特的技术亮点在于强大的网关架构，将控制平面与核心助手分离，确保跨平台通信的流畅性与扩展性。\n\nOpenClaw 非常适合希望构建个性化工作流的技术爱好者、开发者，以及注重隐私保护且不愿被单一生态绑定的普通用户。只要具备基础的终端操作能力（支持 macOS、Linux 及 Windows WSL2），即可通过简单的命令行引导完成部署。如果你渴望拥有一个懂你",349277,3,"2026-04-06T06:32:30",[13,14,15,16],"Agent","开发框架","图像","数据工具","ready",{"id":19,"name":20,"github_repo":21,"description_zh":22,"stars":23,"difficulty_score":10,"last_commit_at":24,"category_tags":25,"status":17},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,"2026-04-05T11:01:52",[14,15,13],{"id":27,"name":28,"github_repo":29,"description_zh":30,"stars":31,"difficulty_score":32,"last_commit_at":33,"category_tags":34,"status":17},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",154349,2,"2026-04-13T23:32:16",[14,13,35],"语言模型",{"id":37,"name":38,"github_repo":39,"description_zh":40,"stars":41,"difficulty_score":32,"last_commit_at":42,"category_tags":43,"status":17},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。",108322,"2026-04-10T11:39:34",[14,15,13],{"id":45,"name":46,"github_repo":47,"description_zh":48,"stars":49,"difficulty_score":32,"last_commit_at":50,"category_tags":51,"status":17},6121,"gemini-cli","google-gemini\u002Fgemini-cli","gemini-cli 是一款由谷歌推出的开源 AI 命令行工具，它将强大的 Gemini 大模型能力直接集成到用户的终端环境中。对于习惯在命令行工作的开发者而言，它提供了一条从输入提示词到获取模型响应的最短路径，无需切换窗口即可享受智能辅助。\n\n这款工具主要解决了开发过程中频繁上下文切换的痛点，让用户能在熟悉的终端界面内直接完成代码理解、生成、调试以及自动化运维任务。无论是查询大型代码库、根据草图生成应用，还是执行复杂的 Git 操作，gemini-cli 都能通过自然语言指令高效处理。\n\n它特别适合广大软件工程师、DevOps 人员及技术研究人员使用。其核心亮点包括支持高达 100 万 token 的超长上下文窗口，具备出色的逻辑推理能力；内置 Google 搜索、文件操作及 Shell 命令执行等实用工具；更独特的是，它支持 MCP（模型上下文协议），允许用户灵活扩展自定义集成，连接如图像生成等外部能力。此外，个人谷歌账号即可享受免费的额度支持，且项目基于 Apache 2.0 协议完全开源，是提升终端工作效率的理想助手。",100752,"2026-04-10T01:20:03",[52,13,15,14],"插件",{"id":54,"name":55,"github_repo":56,"description_zh":57,"stars":58,"difficulty_score":32,"last_commit_at":59,"category_tags":60,"status":17},4721,"markitdown","microsoft\u002Fmarkitdown","MarkItDown 是一款由微软 AutoGen 团队打造的轻量级 Python 工具，专为将各类文件高效转换为 Markdown 格式而设计。它支持 PDF、Word、Excel、PPT、图片（含 OCR）、音频（含语音转录）、HTML 乃至 YouTube 链接等多种格式的解析，能够精准提取文档中的标题、列表、表格和链接等关键结构信息。\n\n在人工智能应用日益普及的今天，大语言模型（LLM）虽擅长处理文本，却难以直接读取复杂的二进制办公文档。MarkItDown 恰好解决了这一痛点，它将非结构化或半结构化的文件转化为模型“原生理解”且 Token 效率极高的 Markdown 格式，成为连接本地文件与 AI 分析 pipeline 的理想桥梁。此外，它还提供了 MCP（模型上下文协议）服务器，可无缝集成到 Claude Desktop 等 LLM 应用中。\n\n这款工具特别适合开发者、数据科学家及 AI 研究人员使用，尤其是那些需要构建文档检索增强生成（RAG）系统、进行批量文本分析或希望让 AI 助手直接“阅读”本地文件的用户。虽然生成的内容也具备一定可读性，但其核心优势在于为机器",93400,"2026-04-06T19:52:38",[52,14],{"id":62,"github_repo":63,"name":64,"description_en":65,"description_zh":66,"ai_summary_zh":67,"readme_en":68,"readme_zh":69,"quickstart_zh":70,"use_case_zh":71,"hero_image_url":72,"owner_login":73,"owner_name":74,"owner_avatar_url":75,"owner_bio":76,"owner_company":77,"owner_location":77,"owner_email":78,"owner_twitter":77,"owner_website":79,"owner_url":80,"languages":81,"stars":121,"forks":122,"last_commit_at":123,"license":124,"difficulty_score":125,"env_os":126,"env_gpu":127,"env_ram":126,"env_deps":128,"category_tags":134,"github_topics":136,"view_count":32,"oss_zip_url":77,"oss_zip_packed_at":77,"status":17,"created_at":151,"updated_at":152,"faqs":153,"releases":183},7356,"determined-ai\u002Fdetermined","determined","Determined is an open-source machine learning platform that simplifies distributed training, hyperparameter tuning, experiment tracking, and resource management. Works with PyTorch and TensorFlow.","Determined 是一个开源的深度学习平台，旨在让机器学习模型的训练与管理变得更加简单高效。它完美兼容 PyTorch 和 TensorFlow 两大主流框架，帮助开发者轻松应对分布式训练、超参数调优、实验追踪及资源管理等复杂挑战。\n\n在深度学习项目中，手动配置多卡并行训练往往门槛较高，云端的 GPU 资源成本也难以控制，同时大量的实验记录容易导致结果难以复现。Determined 通过一体化的解决方案解决了这些痛点：它能自动加速分布式训练以缩短等待时间，利用智能算法自动寻找最佳模型参数，并通过精细的资源调度降低云端算力开销。此外，其内置的实验追踪功能可完整记录代码快照与性能指标，确保研究过程可追溯、可复现。\n\n这款工具特别适合机器学习工程师、数据科学家以及 AI 研究人员使用。无论是希望在本地快速搭建集群的原型开发者，还是需要在 AWS、GCP 或 Kubernetes 上管理大规模训练任务的企业团队，都能从中受益。\n\nDetermined 的独特之处在于其灵活的接入方式：用户既可以通过简单的 Python 类封装快速迁移现有代码，也能利用强大的命令行界面（CLI）和直观的 We","Determined 是一个开源的深度学习平台，旨在让机器学习模型的训练与管理变得更加简单高效。它完美兼容 PyTorch 和 TensorFlow 两大主流框架，帮助开发者轻松应对分布式训练、超参数调优、实验追踪及资源管理等复杂挑战。\n\n在深度学习项目中，手动配置多卡并行训练往往门槛较高，云端的 GPU 资源成本也难以控制，同时大量的实验记录容易导致结果难以复现。Determined 通过一体化的解决方案解决了这些痛点：它能自动加速分布式训练以缩短等待时间，利用智能算法自动寻找最佳模型参数，并通过精细的资源调度降低云端算力开销。此外，其内置的实验追踪功能可完整记录代码快照与性能指标，确保研究过程可追溯、可复现。\n\n这款工具特别适合机器学习工程师、数据科学家以及 AI 研究人员使用。无论是希望在本地快速搭建集群的原型开发者，还是需要在 AWS、GCP 或 Kubernetes 上管理大规模训练任务的企业团队，都能从中受益。\n\nDetermined 的独特之处在于其灵活的接入方式：用户既可以通过简单的 Python 类封装快速迁移现有代码，也能利用强大的命令行界面（CLI）和直观的 Web UI 进行全流程管控。只需简单的 YAML 配置文件，即可定义从资源分配到搜索策略的所有细节，让复杂的深度学习工作流变得井井有条。","\u003Cp align=\"center\">\u003Cimg width=\"400\" src=\"determined-logo.svg\" alt=\"Determined AI Logo\">\u003C\u002Fp>\n\nDetermined is an all-in-one deep learning platform, compatible with PyTorch and TensorFlow.\n\nIt takes care of:\n\n- Distributed training for faster results.\n- Hyperparameter tuning for obtaining the best models.\n- Resource management for cutting cloud GPU costs.\n- Experiment tracking for analysis and reproducibility.\n\n\u003Cbr\u002F>\n\n\u003Cp align=\"center\">\n\u003Cimg alt=\"Features gif\" src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdetermined-ai_determined_readme_4267c20dc061.gif\">\n\u003C\u002Fp>\n\n# How Determined Works\n\nThe main components of Determined are the Python library, the command line interface (CLI), and the Web UI.\n\n## Python Library\n\nUse the Python library to make your existing PyTorch or Tensorflow code compatible with Determined.\n\nYou can do this by organizing your code into one of the class-based APIs:\n\n```python\nfrom determined.pytorch import PyTorchTrial\n\nclass YourExperiment(PyTorchTrial):\n  def __init__(self, context):\n    ...\n```\n\nOr by using just the functions you want, via the Core API:\n\n```python\nimport determined as det\n\nwith det.core.init() as core_context:\n    ...\n```\n\n## Command Line Interface (CLI)\n\nYou can use the CLI to:\n\n- Start a Determined cluster locally:\n\n```\ndet deploy local cluster-up\n```\n\n- Launch Determined on cloud services, such as Amazon Web Services (AWS) or Google Cloud Platform (GCP):\n\n```\ndet deploy aws up\n```\n\n- Train your models:\n\n```bash\ndet experiment create gpt.yaml .\n```\n\nConfigure everything from distributed training to hyperparameter tuning using YAML files:\n\n```yaml\nresources:\n  slots_per_trial: 8\n  priority: 1\nhyperparameters:\n  learning_rate:\n    type: double\n    minval: .0001\n    maxval: 1.0\nsearcher:\n  name: adaptive_asha\n  metric: validation_loss\n  smaller_is_better: true\n```\n\n## Web UI\n\nUse the Web UI to view loss curves, hyperparameter plots, code and configuration snapshots, model registries, cluster utilization, debugging logs, performance profiling reports, and more.\n\n![Web UI](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdetermined-ai_determined_readme_a8cf3f57294b.png)\n\n# Installation\n\nTo install the CLI:\n\n```bash\npip install determined\n```\n\nThen use `det deploy` to start the Determined cluster locally, or on cloud services like AWS and GCP.\n\nFor installation details, visit the the cluster deployment guide for your environment:\n\n- [Local (on-prem)](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Fsetup-cluster\u002Fdeploy-cluster\u002Fon-prem\u002Foverview.html)\n- [AWS](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Fsetup-cluster\u002Fdeploy-cluster\u002Faws\u002Foverview.html)\n- [GCP](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Fsetup-cluster\u002Fdeploy-cluster\u002Fgcp\u002Foverview.html)\n- [Kubernetes](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Fsetup-cluster\u002Fdeploy-cluster\u002Fk8s\u002Foverview.html)\n- [Slurm\u002FPBS](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Fsetup-cluster\u002Fdeploy-cluster\u002Fslurm\u002Foverview.html)\n\n# Examples\nGet familiar with Determined by exploring the 30+ examples in the [examples folder](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Ftree\u002Fmain\u002Fexamples) and the [determined-examples repo](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined-examples).\n\n# Documentation\n\n- [Documentation](https:\u002F\u002Fdocs.determined.ai)\n- [Quick Start Guide](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Fgetting-started.html)\n- Tutorials:\n  - [PyTorch MNIST Tutorial](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Ftutorials\u002Fpytorch-mnist-tutorial.html)\n  - [TensorFlow Keras MNIST Tutorial](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Ftutorials\u002Ftf-mnist-tutorial.html)\n- User Guides:\n  - [Core API](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Fmodel-dev-guide\u002Fapis-howto\u002Fapi-core-ug.html)\n  - [PyTorch API](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Fmodel-dev-guide\u002Fapis-howto\u002Fapi-pytorch-ug.html)\n  - [Keras API](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Fmodel-dev-guide\u002Fapis-howto\u002Fapi-keras-ug.html)\n  - [DeepSpeed API](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Fmodel-dev-guide\u002Fapis-howto\u002Fdeepspeed\u002Foverview.html)\n\n# Community\n\nIf you need help, want to file a bug report, or just want to keep up-to-date\nwith the latest news about Determined, please join the Determined community!\n\n- [Slack](https:\u002F\u002Fdetermined-community.slack.com) is the best place to\n  ask questions about Determined and get support. [Click here to join our Slack](https:\u002F\u002Fdetermined-community.slack.com).\n- You can also follow us on [YouTube](https:\u002F\u002Fwww.youtube.com\u002F@DeterminedAI) and [Twitter](https:\u002F\u002Fwww.twitter.com\u002FDeterminedAI).\n- You can also join the [community mailing list](https:\u002F\u002Fgroups.google.com\u002Fa\u002Fdetermined.ai\u002Fforum\u002F#!forum\u002Fcommunity)\n  to ask questions about the project and receive announcements.\n- To report a bug, [open an issue](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fissues) on GitHub.\n- To report a security issue, email [`security@determined.ai`](mailto:security@determined.ai).\n\n# Contributing\n\n[Contributor's Guide](CONTRIBUTING.md)\n\n# License\n\n[Apache V2](LICENSE)\n","\u003Cp align=\"center\">\u003Cimg width=\"400\" src=\"determined-logo.svg\" alt=\"Determined AI Logo\">\u003C\u002Fp>\n\nDetermined 是一个一体化深度学习平台，兼容 PyTorch 和 TensorFlow。\n\n它负责：\n\n- 分布式训练，以加快结果产出。\n- 超参数调优，以获得最佳模型。\n- 资源管理，以降低云 GPU 成本。\n- 实验跟踪，以便进行分析和复现。\n\n\u003Cbr\u002F>\n\n\u003Cp align=\"center\">\n\u003Cimg alt=\"Features gif\" src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdetermined-ai_determined_readme_4267c20dc061.gif\">\n\u003C\u002Fp>\n\n# Determined 的工作原理\n\nDetermined 的主要组成部分包括 Python 库、命令行界面 (CLI) 和 Web UI。\n\n## Python 库\n\n使用 Python 库可以使您现有的 PyTorch 或 TensorFlow 代码与 Determined 兼容。\n\n您可以通过将代码组织到基于类的 API 中来实现这一点：\n\n```python\nfrom determined.pytorch import PyTorchTrial\n\nclass YourExperiment(PyTorchTrial):\n  def __init__(self, context):\n    ...\n```\n\n或者通过 Core API 使用您需要的函数：\n\n```python\nimport determined as det\n\nwith det.core.init() as core_context:\n    ...\n```\n\n## 命令行界面 (CLI)\n\n您可以使用 CLI 来：\n\n- 在本地启动 Determined 集群：\n\n```\ndet deploy local cluster-up\n```\n\n- 在云服务上部署 Determined，例如 Amazon Web Services (AWS) 或 Google Cloud Platform (GCP)：\n\n```\ndet deploy aws up\n```\n\n- 训练您的模型：\n\n```bash\ndet experiment create gpt.yaml .\n```\n\n使用 YAML 文件配置从分布式训练到超参数调优的所有内容：\n\n```yaml\nresources:\n  slots_per_trial: 8\n  priority: 1\nhyperparameters:\n  learning_rate:\n    type: double\n    minval: .0001\n    maxval: 1.0\nsearcher:\n  name: adaptive_asha\n  metric: validation_loss\n  smaller_is_better: true\n```\n\n## Web UI\n\n使用 Web UI 可以查看损失曲线、超参数图表、代码和配置快照、模型注册表、集群利用率、调试日志、性能剖析报告等。\n\n![Web UI](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdetermined-ai_determined_readme_a8cf3f57294b.png)\n\n# 安装\n\n安装 CLI：\n\n```bash\npip install determined\n```\n\n然后使用 `det deploy` 在本地或 AWS、GCP 等云服务上启动 Determined 集群。\n\n有关安装详情，请访问适用于您环境的集群部署指南：\n\n- [本地（自建）](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Fsetup-cluster\u002Fdeploy-cluster\u002Fon-prem\u002Foverview.html)\n- [AWS](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Fsetup-cluster\u002Fdeploy-cluster\u002Faws\u002Foverview.html)\n- [GCP](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Fsetup-cluster\u002Fdeploy-cluster\u002Fgcp\u002Foverview.html)\n- [Kubernetes](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Fsetup-cluster\u002Fdeploy-cluster\u002Fk8s\u002Foverview.html)\n- [Slurm\u002FPBS](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Fsetup-cluster\u002Fdeploy-cluster\u002Fslurm\u002Foverview.html)\n\n# 示例\n\n通过探索 [examples 文件夹](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Ftree\u002Fmain\u002Fexamples) 和 [determined-examples 仓库](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined-examples) 中的 30 多个示例，熟悉 Determined。\n\n# 文档\n\n- [文档](https:\u002F\u002Fdocs.determined.ai)\n- [快速入门指南](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Fgetting-started.html)\n- 教程：\n  - [PyTorch MNIST 教程](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Ftutorials\u002Fpytorch-mnist-tutorial.html)\n  - [TensorFlow Keras MNIST 教程](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Ftutorials\u002Ftf-mnist-tutorial.html)\n- 用户指南：\n  - [Core API](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Fmodel-dev-guide\u002Fapis-howto\u002Fapi-core-ug.html)\n  - [PyTorch API](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Fmodel-dev-guide\u002Fapis-howto\u002Fapi-pytorch-ug.html)\n  - [Keras API](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Fmodel-dev-guide\u002Fapis-howto\u002Fapi-keras-ug.html)\n  - [DeepSpeed API](https:\u002F\u002Fdocs.determined.ai\u002Flatest\u002Fmodel-dev-guide\u002Fapis-howto\u002Fdeepspeed\u002Foverview.html)\n\n# 社区\n\n如果您需要帮助、想要提交 bug 报告，或者只是想及时了解 Determined 的最新动态，请加入 Determined 社区！\n\n- [Slack](https:\u002F\u002Fdetermined-community.slack.com) 是提问关于 Determined 问题并获得支持的最佳场所。[点击此处加入我们的 Slack](https:\u002F\u002Fdetermined-community.slack.com)。\n- 您也可以在 [YouTube](https:\u002F\u002Fwww.youtube.com\u002F@DeterminedAI) 和 [Twitter](https:\u002F\u002Fwww.twitter.com\u002FDeterminedAI) 上关注我们。\n- 您还可以加入 [社区邮件列表](https:\u002F\u002Fgroups.google.com\u002Fa\u002Fdetermined.ai\u002Fforum\u002F#!forum\u002Fcommunity)，提出关于项目的问题并接收公告。\n- 如需报告 bug，请在 GitHub 上 [打开一个 issue](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fissues)。\n- 如需报告安全问题，请发送邮件至 [`security@determined.ai`](mailto:security@determined.ai)。\n\n# 贡献\n\n[贡献者指南](CONTRIBUTING.md)\n\n# 许可证\n\n[Apache V2](LICENSE)","# Determined 快速上手指南\n\nDetermined 是一个全功能的深度学习平台，兼容 PyTorch 和 TensorFlow。它专注于分布式训练、超参数调优、资源管理及实验追踪，帮助开发者更高效地训练模型并降低成本。\n\n## 环境准备\n\n在开始之前，请确保满足以下前置条件：\n\n*   **操作系统**：Linux (推荐 Ubuntu 18.04+) 或 macOS。Windows 用户建议使用 WSL2 或 Docker Desktop。\n*   **Python 版本**：Python 3.7 - 3.10。\n*   **深度学习框架**：已安装 PyTorch 或 TensorFlow（根据项目需求）。\n*   **容器运行时**：若部署集群，需安装 Docker 和 Docker Compose；若在云环境或 K8s 部署，需配置相应的云厂商 CLI (aws\u002Fgcloud) 或 kubectl。\n*   **网络加速**：国内用户建议配置 pip 镜像源以加速依赖下载：\n    ```bash\n    pip config set global.index-url https:\u002F\u002Fpypi.tuna.tsinghua.edu.cn\u002Fsimple\n    ```\n\n## 安装步骤\n\n### 1. 安装命令行工具 (CLI)\n\n使用 pip 安装 Determined 客户端：\n\n```bash\npip install determined\n```\n\n### 2. 启动本地集群\n\n安装完成后，可使用一条命令在本地启动一个完整的 Determined 集群（基于 Docker）：\n\n```bash\ndet deploy local cluster-up\n```\n\n> **注意**：此命令会自动拉取所需的 Docker 镜像并启动 Master 和 Agent 服务。启动成功后，终端会显示 Web UI 的访问地址（通常为 `http:\u002F\u002Flocalhost:8080`）。\n\n若需部署到云端（如 AWS, GCP）或 Kubernetes 集群，请使用对应的部署命令，例如：\n```bash\ndet deploy aws up\n```\n\n## 基本使用\n\n### 1. 代码适配\n\n将现有的 PyTorch 或 TensorFlow 代码接入 Determined 非常简单。\n\n**方式 A：类式 API (推荐)**\n继承 `PyTorchTrial` 或 `KerasTrial` 类来组织代码：\n\n```python\nfrom determined.pytorch import PyTorchTrial\n\nclass YourExperiment(PyTorchTrial):\n  def __init__(self, context):\n      # 初始化模型、优化器等\n      ...\n  \n  def train_batch(self, batch, epoch_idx, batch_idx):\n      # 定义单步训练逻辑\n      ...\n```\n\n**方式 B：核心 API (Core API)**\n如果不想重构类结构，可以使用函数式接口包裹现有逻辑：\n\n```python\nimport determined as det\n\nwith det.core.init() as core_context:\n    # 在此上下文中运行你的训练代码\n    # core_context 提供日志记录和指标上报功能\n    ...\n```\n\n### 2. 配置实验\n\n创建一个 YAML 配置文件（例如 `gpt.yaml`），定义资源、超参数搜索策略等：\n\n```yaml\nresources:\n  slots_per_trial: 8\n  priority: 1\nhyperparameters:\n  learning_rate:\n    type: double\n    minval: .0001\n    maxval: 1.0\nsearcher:\n  name: adaptive_asha\n  metric: validation_loss\n  smaller_is_better: true\n```\n\n### 3. 运行实验\n\n使用 CLI 提交实验，Determined 将自动处理分布式训练和超参数搜索：\n\n```bash\ndet experiment create gpt.yaml .\n```\n\n### 4. 查看结果\n\n打开浏览器访问 Web UI（本地默认为 `http:\u002F\u002Flocalhost:8080`），你可以：\n*   实时查看 Loss 曲线和指标图表。\n*   对比不同超参数组合的实验效果。\n*   查看代码快照、日志详情及性能分析报告。\n*   管理模型注册表。\n\n---\n更多详细教程和示例代码，请访问 [官方文档](https:\u002F\u002Fdocs.determined.ai) 或 GitHub 上的 [examples 文件夹](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Ftree\u002Fmain\u002Fexamples)。","某电商推荐算法团队正在基于 PyTorch 训练大规模用户行为预测模型，急需在有限预算下快速迭代出最优参数组合。\n\n### 没有 determined 时\n- 分布式训练需手动编写复杂的多卡通信代码，调试耗时且容易出错，导致模型收敛速度缓慢。\n- 超参数调优依赖人工试错或简陋脚本，无法智能搜索最佳组合，往往错过全局最优解。\n- GPU 资源闲置与争抢并存，缺乏统一调度机制，云厂商账单高昂且资源利用率低下。\n- 实验记录散落在本地日志或笔记中，难以复现特定版本的模型效果，团队协作效率极低。\n\n### 使用 determined 后\n- 仅需通过 YAML 配置文件即可一键启动 8 卡分布式训练，determined 自动处理底层通信，训练速度提升数倍。\n- 内置自适应搜索算法（如 Adaptive ASHA）自动探索学习率等关键参数，以更少试验次数找到精度更高的模型。\n- 动态资源管理根据任务优先级自动分配和回收 GPU 插槽，显著降低云端算力成本，避免资源浪费。\n- Web UI 集中展示所有实验的损失曲线、代码快照及性能报告，团队成员可随时对比分析并无缝复现结果。\n\ndetermined 将繁琐的基础设施运维转化为简单的配置工作，让算法工程师能专注于模型创新而非工程杂务。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdetermined-ai_determined_a8cf3f57.png","determined-ai","Determined AI","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002Fdetermined-ai_defb0db9.png","",null,"ai-open-source@hpe.com","https:\u002F\u002Fdetermined.ai","https:\u002F\u002Fgithub.com\u002Fdetermined-ai",[82,86,90,94,98,102,106,110,114,118],{"name":83,"color":84,"percentage":85},"Go","#00ADD8",44.6,{"name":87,"color":88,"percentage":89},"Python","#3572A5",27.9,{"name":91,"color":92,"percentage":93},"TypeScript","#3178c6",24.4,{"name":95,"color":96,"percentage":97},"SCSS","#c6538c",0.8,{"name":99,"color":100,"percentage":101},"Shell","#89e051",0.7,{"name":103,"color":104,"percentage":105},"PLpgSQL","#336790",0.5,{"name":107,"color":108,"percentage":109},"Makefile","#427819",0.4,{"name":111,"color":112,"percentage":113},"HCL","#844FBA",0.3,{"name":115,"color":116,"percentage":117},"Roff","#ecdebe",0.2,{"name":119,"color":120,"percentage":117},"JavaScript","#f1e05a",3218,371,"2026-04-10T20:22:26","Apache-2.0",4,"未说明","需要 GPU 以进行分布式训练和降低成本（具体型号、显存及 CUDA 版本未在 README 中明确说明，但提及支持 AWS\u002FGCP 云 GPU）",{"notes":129,"python":126,"dependencies":130},"该工具是一个全功能的深度学习平台，支持通过 CLI 在本地、AWS、GCP、Kubernetes 或 Slurm\u002FPBS 上部署集群。用户需将代码封装为特定类或使用 Core API 以兼容该平台。实验配置（如分布式训练槽位、超参数搜索策略）通过 YAML 文件管理。具体系统要求需参考官方文档中的集群部署指南。",[131,132,133],"PyTorch","TensorFlow","determined (CLI\u002FLibrary)",[14,16,135],"其他",[137,138,139,140,141,142,143,144,145,146,147,148,149,150],"deep-learning","machine-learning","ml-platform","ml-infrastructure","hyperparameter-optimization","hyperparameter-search","distributed-training","pytorch","tensorflow","hyperparameter-tuning","kubernetes","data-science","mlops","keras","2026-03-27T02:49:30.150509","2026-04-14T12:27:59.454411",[154,159,164,169,174,179],{"id":155,"question_zh":156,"answer_zh":157,"source_url":158},33034,"使用 det deploy local master-up 部署独立 Master 节点失败且容器迅速退出，如何排查？","该问题通常由网络配置或数据库名称解析失败引起。虽然错误信息简短且容器会自动移除，但建议检查以下几点：\n1. 确认是否存在代理设置（http_proxy）干扰了内部网络通信。\n2. 检查 Docker 网络设置，确保 Master 容器能正确解析数据库主机名（如 determined_determined-db_1）。\n3. 尝试通过自定义配置文件明确指定数据库连接信息，例如创建 master-config-test.yaml：\n```\ndb:\n  host: determined_determined-db_1\n  name: determined\n  port: 5432\n  user: postgres\n```\n然后使用命令启动：det deploy local master-up --master-config-path \u002Fpath\u002Fto\u002Fmaster-config-test.yaml","https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fissues\u002F8305",{"id":160,"question_zh":161,"answer_zh":162,"source_url":163},33035,"升级到 Determined 0.19.4+ 后，TensorBoard 事件文件上传到 S3 失败怎么办？","这是一个已知问题，部分修复已合并到主分支，主要解决了 Trial API 内部使用的 TorchWriter 实例关闭问题。但对于用户自行创建的 TorchWriter 实例，目前唯一的解决方案是手动管理其生命周期。\n建议在代码中显式关闭不再使用的 TorchWriter 实例，或者在实验结束后确保正确清理资源。官方计划在未来重新审视 TensorBoard 的支持架构以彻底解决此问题。","https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fissues\u002F5156",{"id":165,"question_zh":166,"answer_zh":167,"source_url":168},33036,"启动 JupyterLab 时设置了 slot=0，为什么在环境中仍然能看到所有 GPU？","该问题已在最新版本中修复。如果仍遇到此情况，通常是因为 Kubernetes 集群中的 NVIDIA GPU 驱动或设备插件配置不当，导致无法正确隔离 GPU 资源。\n建议检查并重新安装以下组件之一：\n1. NVIDIA GPU Operator（推荐用于大多数集群）。\n2. GCP Daemonset（适用于 Google Cloud）：https:\u002F\u002Fraw.githubusercontent.com\u002FGoogleCloudPlatform\u002Fcontainer-engine-accelerators\u002Fmaster\u002Fnvidia-driver-installer\u002Fcos\u002Fdaemonset-preloaded.yaml\n3. NVIDIA Device Plugin：https:\u002F\u002Fraw.githubusercontent.com\u002FNVIDIA\u002Fk8s-device-plugin\u002Fmaster\u002Fnvidia-device-plugin.yml\n确保驱动安装正确后，Kubernetes 才能正确统计和分配 GPU 需求。","https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fissues\u002F3820",{"id":170,"question_zh":171,"answer_zh":172,"source_url":173},33037,"如何在 Web UI 启动 Notebook 时预设默认的挂载路径和配置？","目前可以通过修改 Master 配置文件来设置任务容器的默认行为。在 master config 中添加 task_container_defaults 字段，例如：\n```\ntask_container_defaults:\n  bind_mounts:\n    - host_path: \u002Ftmp\u002Fabc\n      container_path: \u002Ftmp\u002Fc-abc\n    - host_path: \u002Ftmp\u002Fxyz\n      container_path: \u002Ftmp\u002Fc-xyz\n```\n注意：如果在 Web UI 表单中留空直接点击启动，可能会因配置解析问题导致失败。建议先选择一次选项让系统记住配置，或者确保 API 请求中包含完整的配置参数。未来版本计划支持在 UI 下拉菜单中选择多个预设配置。","https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fissues\u002F1583",{"id":175,"question_zh":176,"answer_zh":177,"source_url":178},33038,"在不使用预训练权重的情况下，ImageNet 训练准确率大幅下降如何解决？","当不使用预训练权重从头训练 ImageNet 时，准确率下降是常见现象，尤其是在分布式环境（如多机训练）与单机测试结果不一致时。建议检查以下几点：\n1. 确认多机训练时的学习率是否根据总批次大小（Global Batch Size）进行了正确的线性缩放。\n2. 检查数据加载器（DataLoader）在多机环境下是否正确划分了数据集，避免重复或遗漏。\n3. 对比单机与多机环境的随机种子设置是否一致。\n4. 验证通信后端（如 NCCL）是否工作正常，梯度同步是否有延迟或丢失。","https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fissues\u002F5661",{"id":180,"question_zh":181,"answer_zh":182,"source_url":173},33039,"Web UI 启动 Notebook 时报错，但在编辑完整配置后又能成功，原因是什么？","这通常是因为首次启动时表单字段为空，导致生成的配置缺少必要参数从而引发错误。一旦用户手动选择或填写了选项（即使只是默认值），系统会缓存这些配置并在后续启动时复用，因此不再报错。\n解决方法：\n1. 避免直接点击启动空白表单，先展开“编辑完整配置”确认参数。\n2. 或者在 Master 配置中预先设定 task_container_defaults，确保即使 UI 传参为空也有默认值兜底。\n开发团队已注意到此体验问题并计划改进表单验证逻辑。",[184,189,194,199,204,209,214,219,224,229,234,239,244,249,254,259,264,269,274,279],{"id":185,"version":186,"summary_zh":187,"released_at":188},247778,"v0.38.1","## 发行说明\n[v0.38.1](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fblob\u002Fv0.38.1\u002Fdocs\u002Frelease-notes.rst)\n\n## 更改日志\n* 6bcda5dd3 文档：添加 0.38.1 的发行说明 (#10258)\n* edede259a 杂项：锁定 aiohttp-cors 版本 (#10257)\n* 7f50cfb63 文档：更新依赖项 (#10256)\n* 9ca0efd49 CI：允许发布候选版本触发 Helm Chart 上传 (#10255)\n* 5c2b884aa CI：修复 goreleaser 和 rc 标签中的错误字段键 (#10254)\n* 20d4553a0 文档：在已弃用的文档中添加警告并更新链接 (#10252)\n* 3a8839e59 杂项：将 swagger-ui 升级至 v5.20.0 (#10250)\n* af0a6f781 CI：修复 OSS 和 EE 的发布流程，并新增 EE 干运行发布流程 (#10249)\n* d90dd4cc1 CI：移除对 codecov 的依赖 (#10238)\n* d60d236fe 构建（依赖）：将 golang.org\u002Fx\u002Fcrypto 从 0.24.0 升级至 0.31.0 (#10236)\n* a0ce6767c CI：移除部署最新 GKE 集群的步骤 (#10235)\n* 21f3c737d CI：移除部署最新 main 和 preview 集群的步骤 (#10234)\n* eb7eb8adc CI：修复 apex 安装问题 [DT-5] (#10233)\n* 49eb16601 测试：更新 CI 设置以运行 GPU 单元测试 (#10230)\n* ad01f85fb 杂项：0.38.1 环境镜像 (#10247)\n* 43a9498db 修复：修复 daist 标准输出日志过滤器中的小于\u002F大于符号错误。\n\n","2025-03-19T23:12:44",{"id":190,"version":191,"summary_zh":192,"released_at":193},247779,"v0.38.1-ee","## 发行说明\n[v0.38.1-ee](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fblob\u002Fv0.38.1-ee\u002Fdocs\u002Frelease-notes.rst)\n\n## 更改日志\n* 6bcda5dd3 文档：添加 0.38.1 的发行说明 (#10258)\n* edede259a 杂项：锁定 aiohttp-cors 版本 (#10257)\n* 7f50cfb63 文档：更新依赖项 (#10256)\n* 9ca0efd49 CI：允许发布候选版本触发 Helm Chart 上传 (#10255)\n* 5c2b884aa CI：修复 goreleaser 和 rc 标签中的错误字段键 (#10254)\n* 20d4553a0 文档：在已弃用的文档中添加警告并更新链接 (#10252)\n* 3a8839e59 杂项：将 swagger-ui 升级至 v5.20.0 (#10250)\n* af0a6f781 CI：修复 OSS 和 EE 的发布流程，并新增 EE 干运行发布流程 (#10249)\n* d90dd4cc1 CI：移除对 codecov 的依赖 (#10238)\n* d60d236fe 构建（依赖）：将 golang.org\u002Fx\u002Fcrypto 从 0.24.0 升级至 0.31.0 (#10236)\n* a0ce6767c CI：移除部署最新 GKE 集群的步骤 (#10235)\n* 21f3c737d CI：移除部署最新 main 和 preview 集群的步骤 (#10234)\n* eb7eb8adc CI：修复 apex 安装 [DT-5] (#10233)\n* 49eb16601 测试：更新 CI 设置以运行 GPU 单元测试 (#10230)\n* ad01f85fb 杂项：0.38.1 环境镜像 (#10247)\n* 43a9498db 修复：修复 daist 标准输出日志过滤器中的小于\u002F大于符号错误。\n\n","2025-03-19T23:12:21",{"id":195,"version":196,"summary_zh":197,"released_at":198},247780,"v0.38.0-ee","## 发行说明\n[v0.38.0-ee](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fblob\u002Fv0.38.0-ee\u002Fdocs\u002Frelease-notes.rst)\n\n## 更改日志\n* 0fb554ea4 ci: 修复 0.38.0-ee 版本发布\n* 79ea08a21 ci: 添加 gorelease ee 干运行\n* e3161d653 ci: 移除对 codecov 的依赖 (#10238)\n* e6d195207 ci: 移除部署最新 GKE 集群的步骤 (#10235)\n* f4e590964 ci: 移除部署最新 main 和 preview 集群的步骤 (#10234)\n* 7b1fc8e5d ci: 修复 apex 安装 [DT-5] (#10233)\n* de1c0f11d 测试: 更新 CI 设置以运行 GPU 单元测试 (#10230)\n* 715442484 杂项: 0.38.0 版本发布说明 (#10231)\n* 13e49a7e3 [自动回迁 release-0.38.0] 10226: 杂项: 消除对 fury 仓库的使用 (#10229)\n* a554cd09e [自动回迁 release-0.38.0] 10224: 修复: 使部分 k8s 测试通过 (#10228)\n* 0d373b162 [自动回迁 release-0.38.0] 10221: 修复: 使用新的迁移 gist (#10222)\n* c93b84826 [自动回迁 release-0.38.0] 10213: 修复: 将 k8s 性能修复移植过来 (#10220)\n* 0cc57dfa7 杂项: 将 10208 回迁至 0.38.0 版本 (#10219)\n* 7d9c5ed71 [自动回迁 release-0.38.0] 10216: 修复: 许可证检查测试 (#10217)\n* e2d8f4737 [自动回迁 release-0.38.0] 10206: ci: 从 ci 中移除 datadog (#10214)\n* 9619dcf1f [自动回迁 release-0.38.0] 10211: 杂项: 修复许可证检查 (#10215)\n* 332cefca4 [自动回迁 release-0.38.0] 10207: 修复: 还原: 修复解决无限期排队（STOPPING_COMPLETED）试验的问题 (#10210)\n* e693655b4 [自动回迁 release-0.38.0] 10203: 还原: 日志搜索 (#10205)\n* 50b769048 杂项: 0.38.0 环境镜像 (#10197)\n* bb6f14057 [自动回迁 10160] 修复: maxPoolSlotCapacity 漏洞 (#10195)\n* 7db183ef2 [自动回迁 10182] 文档: 针对移除搜索器上下文所做的文档更改 (#10194)\n* 23f97932c [自动回迁 10192] 修复: Keras 从云端检查点继续训练 (#10193)\n* 508d400f0 [自动回迁 10174] 文档: 更新面向非 Trial 中心世界的文档 (#10186)\n* 87f5ff853 [自动回迁 10188] 修复: 在继续实验配置中包含 max_length (#10190)\n* e72591837 [自动回迁 10183] 文档: 修复发行说明中的错别字 (#10185)\n* 23687dbaa [自动回迁 10178] 文档: tb_plugin 的已知问题 (#10181)\n* 5427a68be [自动回迁 10172] 修复: 禁止在实验\u002F搜索筛选中使用归档列 (#10176)\n* 88c8887c0 [自动回迁 10173] 修复: client.logout() 会重新启用 client.login() (#10177)\n* 42f74e61b [自动回迁 10168] 杂项: 合并自动回迁时忽略 test_e2e_longrunning 测试 (#10179)\n* 020fc4369 [自动回迁 10161] 修复: 修复扩散示例 [DET-10470] (#10169)\n* c69aa6888 [自动回迁 10140] 修复: 设置最大槽位数，且检查点 GC 策略应符合配置策略 (#10167)\n* b5e6315bb 修复: 设置最大槽位数，且检查点 GC 策略应符合配置策略 (#10140)\n* 8e6a65853 [自动回迁 10105] 杂项: 将 det deploy aws 的默认部署类型更改为 simple-rds (#10162)\n* 6fc6710b4 [自动回迁 10153] 文档: 针对配置策略的检查点存储注意事项 (#10165)\n* b366f80da [自动回迁 10138] 功能: 支持 determined_master_host 等 Helm 参数，并提供更好的默认值","2025-03-07T00:27:08",{"id":200,"version":201,"summary_zh":202,"released_at":203},247781,"v0.38.0","## 发行说明\n[v0.38.0](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fblob\u002Fv0.38.0\u002Fdocs\u002Frelease-notes.rst)\n\n## 更改日志\n* 715442484 杂项：发布说明 0.38.0 (#10231)\n* 13e49a7e3 [自动回迁 release-0.38.0] 10226：杂项：移除对 fury 仓库的使用 (#10229)\n* a554cd09e [自动回迁 release-0.38.0] 10224：修复：使部分 k8s 测试通过 (#10228)\n* 0d373b162 [自动回迁 release-0.38.0] 10221：修复：使用新的迁移 gist (#10222)\n* c93b84826 [自动回迁 release-0.38.0] 10213：修复：移植 k8s 性能修复 (#10220)\n* 0cc57dfa7 杂项：将 10208 回迁至 release 0.38.0 (#10219)\n* 7d9c5ed71 [自动回迁 release-0.38.0] 10216：修复：许可证检查测试 (#10217)\n* e2d8f4737 [自动回迁 release-0.38.0] 10206：CI：从 CI 中移除 datadog (#10214)\n* 9619dcf1f [自动回迁 release-0.38.0] 10211：杂项：修复许可证检查 (#10215)\n* 332cefca4 [自动回迁 release-0.38.0] 10207：修复：回滚：修复解决无限期排队（STOPPING_COMPLETED）试验的问题 (#10210)\n* e693655b4 [自动回迁 release-0.38.0] 10203：回滚：日志搜索 (#10205)\n* 50b769048 杂项：0.38.0 环境镜像 (#10197)\n* bb6f14057 [自动回迁 10160] 修复：maxPoolSlotCapacity 漏洞 (#10195)\n* 7db183ef2 [自动回迁 10182] 文档：移除搜索器上下文的相关文档更改 (#10194)\n* 23f97932c [自动回迁 10192] 修复：Keras 从云端检查点继续训练 (#10193)\n* 508d400f0 [自动回迁 10174] 文档：更新非以试验为中心的世界的相关文档 (#10186)\n* 87f5ff853 [自动回迁 10188] 修复：在继续实验配置中包含 max_length (#10190)\n* e72591837 [自动回迁 10183] 文档：修复发行说明中的错别字 (#10185)\n* 23687dbaa [自动回迁 10178] 文档：tb_plugin 的已知问题 (#10181)\n* 5427a68be [自动回迁 10172] 修复：禁止在实验\u002F搜索过滤器中使用归档列 (#10176)\n* 88c8887c0 [自动回迁 10173] 修复：client.logout() 会重新启用 client.login() (#10177)\n* 42f74e61b [自动回迁 10168] 杂项：合并自动回迁时忽略 test_e2e_longrunning 测试 (#10179)\n* 020fc4369 [自动回迁 10161] 修复：修复扩散示例 [DET-10470] (#10169)\n* c69aa6888 [自动回迁 10140] 修复：设置最大槽位数，且检查点 GC 策略应符合配置策略 (#10167)\n* b5e6315bb 修复：设置最大槽位数，且检查点 GC 策略应符合配置策略 (#10140)\n* 8e6a65853 [自动回迁 10105] 杂项：将 det deploy aws 的默认部署类型更改为 simple-rds (#10162)\n* 6fc6710b4 [自动回迁 10153] 文档：关于配置策略的检查点存储注意事项 (#10165)\n* b366f80da [自动回迁 10138] 功能：determined_master_host 等 Helm 支持，更好的默认值 (#10159)\n* d8afc5773 [自动回迁 10155] 修复：修正 iris 示例以使用报告的指标名称 (#10156)\n* 38ae54b67 [自动回迁 10149] 修复：修复重复模型名称的错误信息 (#10154)\n* 47ba6a934 构建：INFENG-943：GoReleaser 配置预发布版本 (#10146)\n* aad58c179 构建：INFENG-942：有条件地绕过 build-react 作业检查 (#10145)\n* d7f0bbfe3 杂项：锁定已发布的 URL 以保持原样","2024-11-22T21:17:18",{"id":205,"version":206,"summary_zh":207,"released_at":208},247782,"0.35.1","## 发行说明\n[0.35.1](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fblob\u002F0.35.1\u002Fdocs\u002Frelease-notes.rst)\n\n## 更改日志\n* 9d4bed2e4 杂项：升级版本号：0.35.1-rc0 -> 0.35.1\n* 46b37617b 修复：修复在列出所有命名空间中的 Pod 时，API 请求过多导致的性能问题（#10202）\n* 5b03599a6 杂项：升级版本号：0.35.0 -> 0.35.1-rc0\n* 4182da4d1 杂项：将当前环境镜像版本升级至 0.35.1\n","2024-11-09T01:04:22",{"id":210,"version":211,"summary_zh":212,"released_at":213},247783,"0.37.0","## 发布说明\n[0.37.0](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fblob\u002F0.37.0\u002Fdocs\u002Frelease-notes.rst)\n\n## 更改日志\n* c41508740 杂项：升级版本：0.37.0-rc4 -> 0.37.0\n* 736fba6fc 文档：添加 0.37.0 的发布说明 (#9995)\n* 73dee980f 文档：修复损坏的链接 (#9996)\n* ecf8ac766 杂项：升级版本：0.37.0-rc3 -> 0.37.0-rc4\n* 1b5030522 修复：修复运行的默认 ID 搜索 (#9988)\n* 0990c117d 杂项：升级版本：0.37.0-rc2 -> 0.37.0-rc3\n* a78b19084 修复：修复 hf on_save 抛出异常的问题 (#9977)\n* 0560939cd 修复：引入 #9963 中的 handleEmptyCell (#9984)\n* 7caf18aff 杂项：升级版本：0.37.0-rc1 -> 0.37.0-rc2\n* 08d782a90 修复：在运行表格中显示搜索进度 (#9976)\n* 478c78fd9 修复：集群页面高度问题 (#9975)\n* 2772a3ce6 修复：修正超参数的 `dataPath` (#9971)\n* 94f2d9514 杂项：升级版本：0.37.0-rc0 -> 0.37.0-rc1\n* 63e7df004 杂项：0.37.0 环境镜像 (#9967)\n* b2267d1a3 杂项：升级版本：0.37.0-dev0 -> 0.37.0-rc0\n* f758303ad 杂项：锁定已发布的 URL 以保留重定向\n* 2a8e7ddca 杂项：锁定 API 状态以便进行向后兼容性检查\n* 3f54d073b 杂项：升级版本：0.36.1-dev0 -> 0.37.0-dev0\n* baf451f20 杂项：对于代理数量为零的资源池，不记录错误日志 (#9960)\n* 6a8606e63 文档：添加 HPC 安装指南 (#9945)\n* 3241edb1d 修复：修复不稳定的一般任务暂停测试 (#9962)\n* 43556e99b 修复：移除用于隐藏 Form.Item 错误信息的 CSS 规则 (#9872)\n* 590600172 性能：提升初始页面加载速度 (#9939)\n* eb1b0de39 文档：添加工作负载告警功能 (#9938)\n* cedfcfe01 杂项：重构并测试 RBAC 配置策略的工作 [CM-530] (#9943)\n* 2d884b9ce 文档：添加集群概览 (#9936)\n* e17d12c4a 功能：发布工作负载告警的更新说明和改进 (#9944)\n* 0db2e3bbd CI：使 slurmcluster 测试更加稳定，希望如此 (#9957)\n* 95f079d4e 功能：添加获取全局配置策略的 API (#9952)\n* d943d852e 杂项：修复任务配置策略的全局 PUT 请求 (#9941)\n* 410edf6a8 修复：修复端到端测试中 MNIST 下载失败的问题 (#9937)\n* 004c194fe CI：修复 test_allocation_csv 测试不稳定的问题 (#9953)\n* 88a4c679b 功能：添加配置策略的 GET API，并修改 CRUD 函数以同时接受两种工作负载类型 (#9946)\n* a73c8db9a 测试：调试身份验证 [TESTENG-95] (#9942)\n* 13db674b5 测试：实验列表显示已归档筛选器 [ET-753] (#9932)\n* 02e302fc8 杂项：从代码编辑器中移除未使用的语言 (#9898)\n* f6d874da1 文档：替换 Slack 链接 (#9919)\n* 26b0954dc 杂项：实现删除配置策略 API 处理程序 (#9927)\n* 2d12be1b8 测试：添加项目测试 [CM-467] (#9928)\n* 062cb52a0 修复：为试验和集群拓扑使用不同的模块 (#9917)\n* 092895818 杂项：更改日志保留策略的日志级别 (#9935)\n* b559467f6 杂项：提高覆盖率目标 (#9920)\n* 3a2ea5629 修复：不要对混合槽类型池过滤槽位 (#9902)\n* a58ed7c3d 杂项：将 RM 代码重新分配给 CM 在 CODEOWNERS 中 (#9926)\n* cb3515e02 修复：当主节点启动或升级时，从主配置中更新 LogRetentionDays (#9930)\n* 13b7b3f02 CI：增加 k8s 集成测试的超时时间 (#992","2024-09-30T15:28:11",{"id":215,"version":216,"summary_zh":217,"released_at":218},247784,"0.36.0","## 发行说明\n[0.36.0](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fblob\u002F0.36.0\u002Fdocs\u002Frelease-notes.rst)\n\n## 更改日志\n* c349314f0 杂项：升级版本：0.36.0-rc7 -> 0.36.0\n* 39db2a81d 文档：添加 0.36.0 的发行说明 (#9854)\n* 61538a203 杂项：升级版本：0.36.0-rc6 -> 0.36.0-rc7\n* 94948238f 修复：修复在 Workpace Creator 视图中弹出的错误提示框 (#9855)\n* bd332285a 杂项：升级版本：0.36.0-rc5 -> 0.36.0-rc6\n* fa155deca 杂项：升级版本：0.36.0-rc4 -> 0.36.0-rc5\n* 9332ab9cc 杂项：0.36.0 环境镜像 (#9851)\n* 838cafeac 撤销“杂项：为部分后端 API 添加追踪信息” (#9843)\n* 1e2447d53 杂项：升级版本：0.36.0-rc3 -> 0.36.0-rc4\n* f70a03d29 修复：更新损坏的 tensorflow 和 certbot 链接 (#9846)\n* e3695a9cc 性能优化：移除 `ExpMetricNames` 接口中重复的 ID (#9848)\n* 101441d83 文档：修复损坏的链接 (#9845)\n* 8e2849306 文档：记录 rbac editorprojectrestricted 角色 (#9844)\n* 9e73cd3b5 杂项：升级版本：0.36.0-rc2 -> 0.36.0-rc3\n* 8acaee55a 杂项：为部分后端 API 添加追踪信息 (#9841)\n* 46a400eb0 修复：将筛选表单在平铺运行视图中的文案改为“显示运行” [ET-740] (#9840)\n* 119d544b5 杂项：升级版本：0.36.0-rc1 -> 0.36.0-rc2\n* 5affb0954 杂项：为 PR 9822 添加发行说明 (#9837)\n* 21bc083f6 功能：Rocm bumpenvs (#9830)\n* 26f8ed22d 杂项：升级版本：0.36.0-rc0 -> 0.36.0-rc1\n* 89d5ddb2c 修复：由于 Docutil 2.0 移除了 rawsource 属性，直接用 node 替代 rawsource 属性 (#9838)\n* d58ff68d5 功能：添加关于 Aurora V1 和 Postgres 12 的 EOL 通知，以及针对 Postgres \u003C=12 的 Master Log 警告 [CM-413] [CM-416] (#9832)\n* 4be07af6a 文档：小幅文档优化 (#9836)\n* 34b567e87 杂项：升级版本：0.36.0-dev0 -> 0.36.0-rc0\n* e11629be5 杂项：锁定已发布的 URL 以保留重定向\n* 6e0b9d1d3 杂项：锁定 API 状态以便进行向后兼容性检查\n* e1a227382 杂项：升级版本：0.35.1-dev0 -> 0.36.0-dev0\n* 42c2efae4 文档：文档清理 (#9834)\n* 3ed0a3973 文档：使文档与以运行为中心的 UX 保持一致 (#9824)\n* a367cd0f0 杂项：弃用自定义搜索器 [MD-504] (#9829)\n* f7846cb9b 功能：允许具有 Viewer 及以上角色的用户查看资源配额 (#9822)\n* 97353c95a 修复：组和用户管理 (CM-436) (#9825)\n* 358ed28a4 修复：如果没有元数据，则隐藏元数据部分 (#9823)\n* 287f3be36 杂项：取消跳过不稳定测试 (#9819)\n* e85ac893a 向 mldm 清晰说明基础数据 lineage (#9828)\n* c0ca6590b 修复：检查点表格操作菜单不应因轮询而消失 [ET-277] (#9812)\n* 740b0e748 文档：描述基本 lineage 步骤 (#9813)\n* e5d4b7f43 杂项：首次支持 k8s rocm [CM-367] (#9794)\n* 9548790e7 杂项：将 torch 版本修复为 2.2.2，适用于 intel mac (#9821)\n* b2a82e896 杂项：弃用带有抢占调度器的 kubernetes 优先级 (#9763)\n* 2002bf02d 文档：获取检查点中的文件列表 (#9818)\n* 91d0b6779 文档：修复损坏的链接 (#9816)\n* e3578490b 修复：不要忽略实验关闭过程中的失败 (#9693)\n* 9b9641631 测试：为实验批量操作添加 go 单元测试 [ET-138] (#965","2024-08-23T21:25:59",{"id":220,"version":221,"summary_zh":222,"released_at":223},247785,"0.35.0","## 发行说明\n[0.35.0](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fblob\u002F0.35.0\u002Fdocs\u002Frelease-notes.rst)\n\n## 更改日志\n* 7d1b0df82 杂项：升级版本：0.35.0-rc20 -> 0.35.0\n* e770ee51f 文档：添加 0.35.0 的发行说明 (#9786)\n* 7f03a87e6 杂项：升级版本：0.35.0-rc19 -> 0.35.0-rc20\n* 3c9a18857 修复：防止在比较视图选择中多次调用时间序列 (#9805)\n* c65c6cd8e 杂项：升级版本：0.35.0-rc18 -> 0.35.0-rc19\n* da58c922e 修复：防止对搜索端点进行额外的初始调用 (#9782)\n* 8074fd97d 杂项：升级版本：0.35.0-rc17 -> 0.35.0-rc18\n* 6fed7664c 杂项：更改 values.yaml 中 defaultNamespace 的注释 (#9793)\n* 6d9b7804f 杂项：升级版本：0.35.0-rc16 -> 0.35.0-rc17\n* f02b6b52f 修复：从链接分叉而来 (#9798)\n* 7928af17f 杂项：升级版本：0.35.0-rc15 -> 0.35.0-rc16\n* 1841a8eb7 修复：不要在比较视图中过滤单个运行 (#9789)\n* 0b89fad5b 杂项：升级版本：0.35.0-rc14 -> 0.35.0-rc15\n* c45148294 修复：取消注释 Helm 值 (#9790)\n* f04144027 杂项：升级版本：0.35.0-rc13 -> 0.35.0-rc14\n* f144957ba 修复：修复了 Helm 图表值和 master-config.yaml (#9788)\n* bfe79123d 杂项：升级版本：0.35.0-rc12 -> 0.35.0-rc13\n* 2794cdcd8 杂项：添加集群名称标题并更改 Helm 值 (#9775)\n* 720dcbbb3 杂项：升级版本：0.35.0-rc11 -> 0.35.0-rc12\n* 94f916d2f 修复：修复超参数和元数据的包含筛选器 (#9779)\n* 501d45c97 杂项：升级版本：0.35.0-rc10 -> 0.35.0-rc11\n* 44e478627 功能：为扁平化运行添加检查点视图 [ET-658] (#9769)\n* e1ff8bb79 杂项：升级版本：0.35.0-rc9 -> 0.35.0-rc10\n* 5314c58c7 功能：在运行页面添加代码选项卡 [ET-657] (#9771)\n* f23ca4c2f 杂项：升级版本：0.35.0-rc8 -> 0.35.0-rc9\n* 2b1f0e703 修复：在运行表格筛选中使用运行检查点数据而非实验数据 (#9767)\n* b750c30a5 修复：从实验负载中提取搜索器指标 (#9768)\n* 461434e7a 杂项：升级版本：0.35.0-rc7 -> 0.35.0-rc8\n* 5cb9a328a 修复：修复恢复分配时缺失的 task_stats start_time (#9745)\n* ca4df7739 杂项：将当前环境镜像版本升级至 0.35.0 (#9760)\n* b0b9d84b3 杂项：将 numpy 版本固定为 1.x [MD-470] (#9748)\n* 375244d1d 撤销“杂项：0.35.0 镜像 (#9732)”\n* bd4af9d4e 杂项：从 RP 描述中移除 RM 名称 (#9758)\n* 22c5ae903 杂项：升级版本：0.35.0-rc6 -> 0.35.0-rc7\n* 2b76ac84d 重构：将 ManageJob 模态框中的“关闭”按钮改为“保存” [DET-10446] (#9746)\n* f739956c8 修复：在搜索视图中为单个运行搜索加载试验数据 (#9742)\n* d1520a429 杂项：升级版本：0.35.0-rc5 -> 0.35.0-rc6\n* ed47fb07a 修复：减少来自 Workspace Create Modal 的 API 调用次数 (#9735)\n* e345871f7 修复：将 FlatRun 协议中的 external_run_id 类型更改为字符串类型 (#9744)\n* 8ef93f888 杂项：微调 CLI 命令关于槽位限制的错误和帮助信息 (#9743)\n* c17fc727f 杂项：添加 ComparisonView 修复的发行说明 (#9741)\n* 1dbdf003b 杂项：升级版本：0.35.0-rc4 -> 0.35.0-rc5\n* 27b8dbd92 修复：死锁问题 (#9728)\n* f939bc49f 杂项：升级版本","2024-08-08T18:10:22",{"id":225,"version":226,"summary_zh":227,"released_at":228},247786,"0.34.0","## 发行说明\n[0.34.0](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fblob\u002F0.34.0\u002Fdocs\u002Frelease-notes.rst)\n\n## 更改日志\n* ede239684 杂项：版本号更新：0.34.0-rc12 -> 0.34.0\n* f0d825da9 杂项：版本号更新：0.34.0-rc11 -> 0.34.0-rc12\n* 1556c181e 修复：暂停\u002F恢复运行测试不稳定问题 (#9592)\n* a74e389be 文档：添加 0.34.0 的发行说明 (#9561)\n* e5fc5f1a4 杂项：版本号更新：0.34.0-rc10 -> 0.34.0-rc11\n* a51a640f1 修复：工作空间中项目的编辑\u002F移动模态框意外关闭 [DET-10388] (#9588)\n* ce3ea17ff 杂项：版本号更新：0.34.0-rc9 -> 0.34.0-rc10\n* 5b40a5c08 杂项：移除 Circle CI 的共享集群测试 (#9579)\n* 0b4dec468 杂项：版本号更新：0.34.0-rc8 -> 0.34.0-rc9\n* 01baf33dc 杂项：发布 0.34.0 bumpenvs (#9578)\n* bad22b2bd 杂项：添加 NVIDIA 驱动版本匹配测试，并更新环境变量 [MD-413] (#9567)\n* 9adbe7c14 撤销“杂项：0.34.0 bumpenvs (#9565)”\n* 60ada0c5f 杂项：版本号更新：0.34.0-rc7 -> 0.34.0-rc8\n* cde8a186a 修复：笔记本空闲状态负载错误 [MD-447] (#9571)\n* 3f292f59b 杂项：版本号更新：0.34.0-rc6 -> 0.34.0-rc7\n* f36b11065 修复：修正 allocation_workspace_info 表中的 workspace_id 列类型 (#9574)\n* f0f45f860 杂项：版本号更新：0.34.0-rc5 -> 0.34.0-rc6\n* a6c7918c7 修复：为历史分配持久化保存工作空间 ID\u002F名称及实验 ID [DET-10378] (#9550)\n* f66e81634 杂项：版本号更新：0.34.0-rc4 -> 0.34.0-rc5\n* eaabab17b 修复：为项目密钥的补丁操作添加验证 (ET-305)\n* d8b80ad87 修复：不修改缓存的 GetAgentsResponse (#9569)\n* 25804d70d 杂项：版本号更新：0.34.0-rc3 -> 0.34.0-rc4\n* ead223225 修复：在项目详情页面的面包屑导航中返回工作空间名称 (#9564)\n* 8da67d2ee 杂项：0.34.0 bumpenvs (#9565)\n* 2677dc2f4 杂项：版本号更新：0.34.0-rc2 -> 0.34.0-rc3\n* 42bea1ab5 杂项：修复 boto3 依赖项语法 (#9551)\n* b2c7e22d9 杂项：版本号更新：0.34.0-rc1 -> 0.34.0-rc2\n* 2f1283d6a 修复：仅在弱密码确实适用时才显示警告 [DET-10216] (#9538)\n* ca208b92c 杂项：版本号更新：0.34.0-rc0 -> 0.34.0-rc1\n* f15bda886 功能：det deploy local 现可为您生成密码 [DET-10197] (#9518)\n* abaf2e349 杂项：版本号更新：0.34.0-dev0 -> 0.34.0-rc0\n* 0cf7aba52 杂项：锁定已发布的 URL，以保留重定向\n* cd85b4441 杂项：锁定 API 状态，以便进行向后兼容性检查\n* 25b629996 杂项：版本号更新：0.33.1-dev0 -> 0.34.0-dev0\n* 83b9a8b17 功能：为笔记本和 Shell 任务添加连接模态框 [MD-404] (#9545)\n* b9ea173f7 杂项：Bumpenvs 8c90e80 (#9544)\n* f9a5dd515 修复：更新 getProjectColumns 调用 (ET-270) (#9509)\n* 325d47e9a 修复预提交 Lint 检查 (#9543)\n* 553521e84 功能：为 Jupyter 笔记本启用令牌认证 [MD-404] (#9452)\n* ea929fc69 测试：det 框架支持“第 n 个”组件 [testeng-1] (#9540)\n* 756812988 文档：解决两个链接检查失败问题 (#9539)\n* 3641bfcea 功能：支持在远程 k8s 集群上代理运行 Determined 任务 (#9469)\n* 44f446c26 修复：Huggingface Trust Remote Repo (#9535)\n* 332010756 杂项：允许空运行元数据请求删除现有内容","2024-06-28T20:09:57",{"id":230,"version":231,"summary_zh":232,"released_at":233},247787,"0.33.0","## 发行说明\n[0.33.0](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fblob\u002F0.33.0\u002Fdocs\u002Frelease-notes.rst)\n\n## 更改日志\n* 0c2d3cfe9 杂项：升级版本：0.33.0-rc5 -> 0.33.0\n* e16554188 文档：添加 0.33.0 的发行说明 (#9444)\n* 8c69d8bc0 杂项：升级版本：0.33.0-rc4 -> 0.33.0-rc5\n* e1a40b127 修复：在常规 AWS 部署中不使用默认的 EFS 挂载 (#9437)\n* ebe26988b 杂项：升级版本：0.33.0-rc3 -> 0.33.0-rc4\n* b85b8b3fa 修复：正确设置 genai 中 shared_fs 挂载的默认值 (#9433)\n* 52c7d9512 杂项：升级版本：0.33.0-rc2 -> 0.33.0-rc3\n* 9968dcec8 修复：添加用于检查管理员\u002FDetermined 密码是否为空的功能门控 [DET-10197] (#9425)\n* 9c4fd74c7 杂项：升级版本：0.33.0-rc1 -> 0.33.0-rc2\n* 274b152f4 修复：当配置无效时保持模板模态框打开 (#9424)\n* f4d6f54b8 杂项：升级版本：0.33.0-rc0 -> 0.33.0-rc1\n* 2661ae06d 杂项：为发布更新 NGC 镜像版本 (#9418)\n* cbc15db71 修复：主节点在迁移之前先检查数据库的新旧程度 [DET-10312] (#9414)\n* d1b3343d1 杂项：升级版本：0.33.0-dev0 -> 0.33.0-rc0\n* ca451981f 杂项：锁定已发布的 URL 以保留重定向\n* f2cd0181e 杂项：锁定 API 状态以进行向后兼容性检查\n* 6184f6fa3 杂项：升级版本：0.32.1-dev0 -> 0.33.0-dev0\n* 4af9bfc39 撤销：框架拆分 (#9405)\n* 6fa1420d4 测试：项目创建和删除的 React 端到端测试 [INFENG-456] (#9244)\n* 860f6a8da 文档：描述 WebUI 配置模板 (#9399)\n* 6ff8eb712 杂项：添加 Slurm 代码所有者 (#9403)\n* 68b36c6fe 功能：要求在新集群启动时设置初始密码 [DET-10197] (#9314)\n* 0ef3e104d 测试：数据网格滚动 [INFENG-687] (#9379)\n* 18ee0e31d 杂项：更新 Docker 重新标记脚本 (#9401)\n* 6ed297699 在模型中心测试中固定 setuptools 版本 (#9402)\n* c4ebe5e54 功能：发布带有注释的 WebUI 模板 (#9383)\n* 3bbb51a7f 功能：在“日志”选项卡中显示日志保留天数和剩余日志保留天数 (#9305)\n* 047580c0e 功能：将默认调度器更新为优先级，用于 agentrm (#9385)\n* ce70c005e 文档：增加 Helm 安装密码的相关信息 (#9388)\n* b84ee1f00 文档：改进集群可观性文档和仪表盘 (#9391)\n* c3b3ae6ef 功能：Helm 安装会检查密码复杂度 [DET-10293] (#9360)\n* 5c51164b3 修复：跳过对非托管实验的资源检查 (#9372)\n* 107e10851 功能：在“平铺运行”视图中添加排序菜单 (#9396)\n* cb81a44e0 功能：在比较视图中添加图表 (ET-99) (#9215)\n* cd33c13ff 测试：将易出错的修复恢复回来 [INFENG-694] (#9394)\n* d3e89b108 文档：为第二个非托管示例添加实验配置。(#9397)\n* d4e23f4d5 杂项：将 requests 版本固定在 \u003C 2.32.0，以便 Docker 正常工作 (#9395)\n* 5480c57db 杂项：不要为 views_and_triggers 使用单独的模式 (#9392)\n* 893f7f5c0 杂项：添加 resource_pools 集成测试 (#9356)\n* de215938b 杂项：按每次提交推送 OSS 镜像 (#9386)\n* 95c70d48f 文档：为 genai 文档添加导航 (#9387)\n* 0c42ced90 功能：SDK 方法用于获取 Pachyderm 配置 [MD-406] (#9348)\n* 0ff09e038 文档：描述 WebUI 的密码要求 (#9378)\n* 31bc08","2024-05-29T20:46:58",{"id":235,"version":236,"summary_zh":237,"released_at":238},247788,"0.32.1","## Release Notes\n[0.32.1](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fblob\u002F0.32.1\u002Fdocs\u002Frelease-notes.rst)\n\n## Changelog\n* 7d0b38a74 chore: bump version: 0.32.1-rc0 -> 0.32.1\n* 351826c9b docs: add release notes for 0.32.1 (#9351)\n* 947585fee chore: bump version: 0.32.1-dev0 -> 0.32.1-rc0\n* f9da12f89 chore: lock api state for backward compatibility check\n* 1e8f8de21 fix: remove pod labels with potentially incompatible names (#9349)\n* 6995ca677 chore: bump version: 0.32.0 -> 0.32.1-dev0\n\n","2024-05-10T17:24:08",{"id":240,"version":241,"summary_zh":242,"released_at":243},247789,"0.32.0","## Release Notes\n[0.32.0](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fblob\u002F0.32.0\u002Fdocs\u002Frelease-notes.rst)\n\n## Changelog\n* a1b724210 chore: bump version: 0.32.0-rc8 -> 0.32.0\n* d8580c227 docs: add release notes for 0.32.0 (#9301)\n* 2244f71ef chore: bump version: 0.32.0-rc7 -> 0.32.0-rc8\n* 0322dc799 fix: filter action experiments, old ExperimentList (#9325)\n* 5ebb00861 chore: bump version: 0.32.0-rc6 -> 0.32.0-rc7\n* b2087945e fix: filter batch action experiments (#9316)\n* 991818b75 chore: bump version: 0.32.0-rc5 -> 0.32.0-rc6\n* e2777824d fix: Bulk Action bug (#9255)\n* b2663af49 chore: bump version: 0.32.0-rc4 -> 0.32.0-rc5\n* ee63b679a fix: users can be removed from all groups in Web UI (#9259)\n* 00b95c361 chore: bump version: 0.32.0-rc3 -> 0.32.0-rc4\n* 642e32354 fix: historical-usage date calculation bug (#9257)\n* f50698972 chore: bump version: 0.32.0-rc2 -> 0.32.0-rc3\n* 1047e7810 fix: hew update for select bug in log viewer (#9249)\n* 4c59c9c64 chore: bump version: 0.32.0-rc1 -> 0.32.0-rc2\n* f8ad00933 fix: undo default log retention in values.yaml (#9245)\n* 4b3adb952 docs: add a release note for aurora issue. (#9241)\n* 004fe7022 fix: allow genai deployments with agent GIDs set to share data properly (#9243)\n* be231d9fd chore: bump version: 0.32.0-rc0 -> 0.32.0-rc1\n* 714264ed4 chore: bump version: 0.32.0-dev0 -> 0.32.0-rc0\n* dc88b9f1e chore: bump version: 0.31.1-dev0 -> 0.32.0-dev0\n* 7ffdadfe5 ci: add determined-ee context to python ee publish (#9234)\n* c18ac836c fix: properly merge resource configs (#9233)\n* 3b39d3cd1 chore: add log retention to help charts (#9216)\n* 36463952e chore: lock published urls to preserve redirects\n* 80d890936 chore: lock api state for backward compatibility check\n* 39b948c09 feat: add genai user role to rbac (#9206)\n* 43289e9af test: ee and oss have separate handling (#9218)\n* 1ca36134b fix: debounce `userSettings` update (#9220)\n* ab382b425 chore: update the license date (#9225)\n* ff10ac0ad docs: Fix broken links (#9219)\n* ac68df87b chore: default observability.enable_prometheus to true (#9222)\n* 26c1940eb chore: upgrade protoc used in CI (#7935)\n* 9f6bbc906 chore: Add streaming updates feature flag [MD-371] (#9190)\n* f8b373617 ci: Exclude deploy\u002FREADME.md from build (#9211)\n* 3bfc212b6 fix: hew update for chart scroll bug (#9210)\n* da8a040fc feat: CLI allows and requires creating a user with a password DET-10184 (#9112)\n* fbccaf18b chore: clean up rm module [RM-202] (#9191)\n* 8caf3cb73 test: user tests [INFENG-455] (#9152)\n* 3568f27a3 fix: Skip expected error from web socket (#9194)\n* 1b212ae38 feat: add kill run endpoint (#9061)\n* e7d870e90 test: use devcluster for react tests [INFENG-449] (#9185)\n* bd4a54e29 fix: shared cluster test to work in OSS again (#9195)\n* b874acb29 docs: fix another instance of broken docs link (#9208)\n* 86be18a83 ci: pass ee into args to prevent latest main deploying as ee (#9207)\n* f74ab9ceb docs: Describe multi rm k8s (#9025)\n* 6fb1c5283 ci: deploy awscli to system (#9188)\n* 9cfbb598a docs: fix nvidia device plugin link for EKS (#9204)\n* 3e865c643 test: skip flakey user provision tests (#9203)\n* 598784d1b chore: make multi-RM an EE-only feature [RM-166] (#9192)\n* 6d2be5216 ci: fix test-det-deploy-local (#9196)\n* 5f312ed53 test: can't launch NSC test assert 404 instead of 403 (#9197)\n* 4b1c9379e test: fix a test util issue with master config schema assumptions (#9193)\n* 0bc13d8d1 feat: non-blocking metrics reports [MD-144] (#9107)\n* 2ced9b9b1 ci: do dry runs of `publish-docs` for RCs (#9186)\n* 72344e083 feat: Use feature flag for streaming updates - manually update project store  (#9170)\n* dd7f4b507 docs: add profiling section for trainer API UG [MD-373] (#9177)\n* 06586f0a4 fix: better exception handling in detached mode (#9183)\n* 283daab2d feat: Unfork Enterprise Edition (EE) and require license key for EE features (#9168)\n* f233c955c docs: FAQ for python SDK ckpt download, k8s deprecation labels. (#9187)\n* 6fcefac79 chore: bump version: 0.31.0-dev0 -> 0.31.1-dev0\n* 19688a98b chore: add docs dropdown link for new version\n* 2b2e96a70 docs: add release notes for 0.31.0 (#9159)\n* b194686e8 chore: style fix for helm initialUserPassword (#9158)\n* a5e9f0c80 chore: add option to auto pick the only matching name on partial hits (#9108)\n* 371c90bfd fix: louden server errors coming from deleteCheckpoints (#9184)\n* 0765e380b chore: pass correct master scheme to genai (#9181)\n* 26f5e0b67 fix: report errors from deletecheckpoints endpoint + improve feedback (#9178)\n* 1037d83ae chore: bumpenv update NGC base images version to 24.03 (#9132)\n* 1cc9cd7f0 fix: count determined-system pods as det pods [RM-148] (#9148)\n* 0fc247cf8 fix: single-searcher MNIST example runs for multiple epochs (#9160)\n* d41c4a705 fix: fix docs and wording (#9179)\n* 5541e540b feat: RM-130 add determined info as pod labels (#9140)\n* ee15da01e test: Djanicek\u002Finfeng 456\u002Fworkspaces and projects (#9117)\n* e6c0c9971 chore: add typing annotations for zmq (#9176)\n* 4ceaed051 docs: Add readme to toc","2024-05-08T15:35:57",{"id":245,"version":246,"summary_zh":247,"released_at":248},247790,"0.31.0","## Release Notes\n[0.31.0](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fblob\u002F0.31.0\u002Fdocs\u002Frelease-notes.rst)\n\n## Changelog\n* 583e0c3d3 chore: bump version: 0.31.0-rc7 -> 0.31.0\n* 29574c4f1 docs: add release notes for 0.31.0 (#9159)\n* 40c34cbcc chore: bump version: 0.31.0-rc6 -> 0.31.0-rc7\n* 75b7e43cc fix: louden server errors coming from deleteCheckpoints (#9184)\n* 44503bb4f chore: bump version: 0.31.0-rc5 -> 0.31.0-rc6\n* 956df4059 fix: fix docs and wording (#9179)\n* dae548dc6 fix: report errors from deletecheckpoints endpoint + improve feedback (#9178)\n* 592280d4a chore: style fix for helm tls (#9163)\n* 7565447bb chore: bump version: 0.31.0-rc4 -> 0.31.0-rc5\n* 2daa1fc54 fix: TensorBoard visualization from batch actions. (#9156)\n* 4bbc20de8 fix: fix disable button condition in launch jupyter notebook modal (#9155)\n* ac15a8623 chore: bump version: 0.31.0-rc3 -> 0.31.0-rc4\n* 990fbfb1e feat: add helm master level config for tcd startup hooks (#9135)\n* a6ae2aa4b ci: publish-docs installs awscli into user space (#9153)\n* d5deffbd1 chore: bump version: 0.31.0-rc2 -> 0.31.0-rc3\n* 7f7d2bfa9 fix: fix docs for log retention (#9149)\n* 776c5c3ac fix: ensure all columns have widths (#9136)\n* e8b4fd712 chore: bump version: 0.31.0-rc1 -> 0.31.0-rc2\n* 691d190cd test: fix test_logging typehint syntax error (#9142)\n* 61999f435 docs: revert helm values change for multirm (#9145)\n* be36ecdb6 chore: bump version: 0.31.0-rc0 -> 0.31.0-rc1\n* f78ccf8d2 docs: revert-multiRM-mc-doc (#9144)\n* 3014dba9e chore: bump version: 0.31.0-dev0 -> 0.31.0-rc0\n* 828532afb chore: lock api state for backward compatibility check\n* 0547f7f78 chore: bump version: 0.30.1-dev0 -> 0.31.0-dev0\n* 55ef649b6 chore: change multirm log messages to trace level [RM-151] (#9138)\n* 1bb2fe496 feat: expose `hyperparameters` in experiments api to avoid using deprecated `config` property for experiment (#9012)\n* 5a588e0f0 chore: lock published urls to preserve redirects\n* f46bc6916 chore: lock api state for backward compatibility check\n* 28b3affc5 feat: add cluster wide startup hook for tasks (#9124)\n* fe2b61666 docs: Describe pwds default accounts (#9137)\n* d1c268b89 fix: down migrations (#9133)\n* f7a526045 chore: update PyPi metadata (#8971)\n* 2c3ce2934 chore: set the default db storage as docker volume instead of a mount (#9127)\n* 8b11e3ae4 ci: publish docs without installing awscli (#9126)\n* 133f83835 fix: prevent table breaking on null columnWidths [ET-161] (#9131)\n* ec438095c fix: det gcp down doesn't have a det_version argument (#9121)\n* c89b3dfee fix: reduce time and increase reliability of tests (#9125)\n* d5f807da0 feat: helm deploys with a password (#9113)\n* 8a7832a6e fix: unlock mutex for experiment ResourcePool() [RM-152] (#9119)\n* e70d38e12 docs: add a python sdk example for log following. (#8981)\n* 3028efb04 docs: add helm doc updates (#9122)\n* cf2f2be52 fix: fix regression caused by join on trials view (#9091)\n* bdab9e47e feat: create Searches view (#9089)\n* 65339d21d chore: PR template again [INFENG-600] (#9118)\n* 0c6985b97 chore: update github PR template [INFENG-600] (#9098)\n* f4b04716d docs: add instructions on deploying determined via HPE MLDES [SAAS-1877] (#9105)\n* c32ac6f49 chore: add test to CODEOWNERS [INFENG-605] (#9115)\n* 25767b9ce fix: helm value for gke tests (#9114)\n* a0847b89b fix: match GetJobQueueStats behavior in k8s RM to agent RM [RM-136] (#9097)\n* 2ef5ab90b chore: better k8s testing with shared gke cluster (#9074)\n* 7fc8d7a06 chore: add nightly gke cluster cleanup job (#9031)\n* 5cb792738 chore: bump version: 0.30.0-dev0 -> 0.30.1-dev0\n* 7f6577945 chore: add docs dropdown link for new version\n* 4ae4075fb docs: add release notes for 0.30.0 (#9103)\n* 9dce6f03b feat: Add model version streaming (#9029)\n* 2c6fec7fb test: user-page-models (#9084)\n* 75b1ff45e feat: det deploy aws adds tags to dynamic agents [RM-140] (#9106)\n* d6059e974 feat: Create MoveRun endpoint (#9001)\n* 91d7e0844 feat: Pre-select ws when launching notebook (#9109)\n* f03a8a8f5 fix: add missing k8s job submission times to allocations (#9028)\n* b8bf3960d chore: upgrade Bun to fix race condition in tests [DET-10193] (#9082)\n* bc8c31cab fix: make sure that the genai helm chart services work across namespaces (#9102)\n* 58cd22b44 ci: INFENG-600 remove single commit legacy validation (#9104)\n* 943b2cb97 fix: prevent checkpoint modals from closing on their own [ET-116] [ET-120] (#9094)\n* d4eed0ed1 chore: change RM log message back to Debug level (#9093)\n* 2af21eebc chore: unshadow more builtins (#9092)\n* 519d702ec docs: update multiRM docs (#9050)\n* 3688c3fc9 fix: job queue panic for multirm [RM-123] (#9079)\n* f78b9aa68 fix: add change in master config to devcluster.yaml (#9087)\n* 8f02a7f02 fix: fix master config and experiment config for log retention (#9075)\n* fe1a6bb82 fix: no more shadowing \"license\" (#9085)\n* 3f2d6abfe fix: spacing issue with exp list pagination (#9067)\n* 37abc6ce5 fix: stop showing loading indicator in `queued` state (#9081)\n* 4050eda43 chore: bumpenvs for","2024-04-17T22:35:51",{"id":250,"version":251,"summary_zh":252,"released_at":253},247791,"0.30.0","## Release Notes\n[0.30.0](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fblob\u002F0.30.0\u002Fdocs\u002Frelease-notes.rst)\n\n## Changelog\n* 5a635188f chore: bump version: 0.30.0-rc5 -> 0.30.0\n* 97aaa02c6 docs: add release notes for 0.30.0 (#9103)\n* c10844310 chore: bump version: 0.30.0-rc4 -> 0.30.0-rc5\n* 4ce78b202 fix: prevent checkpoint modals from closing on their own [ET-116] [ET-120] (#9094)\n* 8bcdcc8dc chore: bump version: 0.30.0-rc3 -> 0.30.0-rc4\n* e90238a6d chore: bump version: 0.30.0-rc2 -> 0.30.0-rc3\n* b8db2e683 fix: slot stats are not filled in everywhere (#9070)\n* d2e3a5c18 fix: remove parent_id from create_experiment (#9068)\n* 61958ef8d fix: API migration to improve performance in resource pool page (#9056)\n* 62d102b06 chore: bump version: 0.30.0-rc1 -> 0.30.0-rc2\n* bc241b6fa docs: Update release notes (#9044)\n* 2e31ece13 fix: loading experiments without filterset (#9059)\n* 4efaede76 chore: bump version: 0.30.0-rc0 -> 0.30.0-rc1\n* d2949d3cc faster migrations (#9060)\n* 4c6e35c56 feat: add slot stats to \u002Fagents endpoints (#9048)\n* f32dc823b chore: bump version: 0.30.0-dev0 -> 0.30.0-rc0\n* 10030a6c0 chore: lock published urls to preserve redirects\n* 220f82067 chore: bump version: 0.29.2-dev0 -> 0.30.0-dev0\n* 1e6f0f7c6 feat: Use filtered resource pools when creating notebook (#9045)\n* 74fe16bfb feat: profiling v2 [MD-27] (#9032)\n* 133d127ba docs: revert multirm docs changes #9016\n* 1992c9786 chore: optional DB migrations (#9047)\n* 84ba68877 fix: docs lint (#9052)\n* 848b216cf feat: add command det model delete (#9039)\n* 1202d5c7c refactor: DET-9976 remove agentID type from agentrm (#9040)\n* 0710c58c7 docs: Describe editorrestricted (#9049)\n* 02da36f36 chore: mark db-dependent tests as needing to run in integration (#9041)\n* 6c88e8dbd fix: move experiment SQL error (#9042)\n* 3fa0df156 Revert \"docs: add EditorRestricted role release note (#9007)\" (#9046)\n* 60cb00380 test: Jcom\u002Finfeng 454\u002Fsign in tests (#9013)\n* f08b40602 ci: tag CI-deployed resources (#9043)\n* 1868723db build(deps): bump google.golang.org\u002Fprotobuf from 1.28.0 to 1.33.0 (#8996)\n* d4ab20bd5 build(deps): bump github.com\u002Fdocker\u002Fdocker (#9026)\n* e4bc377c3 test: playwright config and browser usability (#9024)\n* f6b9ac845 build(deps): bump github.com\u002Fjackc\u002Fpgx\u002Fv4 from 4.12.0 to 4.18.2 (#8987)\n* c811947b6 chore: helm for multirm kubeconfig_path (#9033)\n* 4441d6d4c feat: Add template to py sdk `create_experiment` (#8927)\n* 5ac1b853c chore: revert helm for multirm kubeconfig_path (#9030)\n* 6fec24d66 chore: helm for multirm kubeconfig_path (#9015)\n* 0518785be feat: streaming update code generation for typescript (#8988)\n* 39afa3c5c docs: add documentation for multirm (#9016)\n* 7e37c226b chore: add grpc based auth fallback to proxied requests (#8980)\n* 5e1f2af91 fix: Experiment.await_first_trial exits when Experiment is terminal (#9022)\n* a603f4cc4 chore: logins return Sessions (#8883)\n* 93b6aa295 feat: SearchFlatRuns api call for flat runs table support (#8852)\n* fa43bffdc ci: test-perf uses determined version from github (#9019)\n* 137bfcd9a feat: add model streaming (#8973)\n* 8bf280de0 refactor: consolidate experiment list selection state (#8860)\n* 674cd7302 ci: DRY skip logic and clarity on step name (#9002)\n* 00d145f8b chore: bump version: 0.29.1-dev0 -> 0.29.2-dev0\n* a3ba9e9e4 chore: add docs dropdown link for new version\n* e922a414f docs: add release notes for 0.29.1 (#9014)\n* dfed63d6e chore: reassign ml-sys CODEOWNERship to model-dev (#9000)\n* eac7ddffe test: document ui e2e with backend test instructions for local (#9005)\n* bc1b43173 docs: add EditorRestricted role release note (#9007)\n* f52f43bf3 chore: warn about det deploy det-version mistmach (#8994)\n* 5b17df323 chore: limit code coverage report to files in src; omit generated files (#9003)\n* f73fd092e fix: escape regex in `ProjectDeleteModal` (#8998)\n* 73fd1cdfe feat: Add multi RM name to K8s (#8993)\n* 978a02e2b ci: Djanicek\u002Finfraeng 487\u002Fcircle test runner (#8977)\n* 4730d768b chore: ban http.Transport & http.Client; add cleanhttp (#8991)\n* 52572d437 fix: improved textcell performance for novels (#8986)\n* 89d470824 docs: add EditorRestricted role to rbac docs (#8984)\n\n","2024-04-04T18:24:03",{"id":255,"version":256,"summary_zh":257,"released_at":258},247792,"0.29.1","## Release Notes\n[0.29.1](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fblob\u002F0.29.1\u002Fdocs\u002Frelease-notes.rst)\n\n## Changelog\n* 6f0810b25 chore: bump version: 0.29.1-rc2 -> 0.29.1\n* d13dfac2b docs: add release notes for 0.29.1 (#9014)\n* 8093bee50 chore: bump version: 0.29.1-rc1 -> 0.29.1-rc2\n* cce4e6bb4 chore: warn about det deploy det-version mistmach (#8994)\n* a2576bec3 chore: bump version: 0.29.1-rc0 -> 0.29.1-rc1\n* 05a75b32a fix: escape regex in `ProjectDeleteModal` (#8998)\n* de8d02db5 chore: bump version: 0.29.1-dev0 -> 0.29.1-rc0\n* 0a2fd2855 chore: change GKE version (#8989)\n* 055dd8381 docs: Update Deploy on GCP (#8985)\n* 47cb6fdd1 fix: remove error text in continue trial modal (#8923)\n* 2f40476c8 chore: bump version: 0.29.0-dev0 -> 0.29.1-dev0\n* 84a846e74 chore: add docs dropdown link for new version\n* 0fd6b610f docs: add release notes for 0.29.0 (#8955)\n* 115bf1301 fix: remove duplicate permissions in rbac CLI output (#8972)\n* cc2e9b4ff chore: Bumpenvs for NGC+ images (#8975)\n* 1a35e5d4a test: add e2e_tests for multirm k8s [RM-11] (#8926)\n* 18154f65a chore: add type ResourcePoolName string (#8978)\n* a22656d8b chore: remove panics from rm initialization (#8983)\n* 4ae09875a chore: amend contributing doc to point to correct make rule, as of #2892 (#8947)\n* 26c985c29 fix: Check auth validity before setting isAuthenticated (#8967)\n* f67c47350 fix: nil deref in ReadPreemptionStatus (#8979)\n* 6e1acf4f7 chore: multirm unique resource pool config changes [RM-74] (#8974)\n* ca2987939 chore: add multirm router layer to rm module (#8963)\n* 2395dcb65 fix: stopping states are not handled in restore properly [RM-69] (#8958)\n* b06c92347  feat: allow k8srm to connect with a kubeconfig (#8953)\n* f8f860d48 chore: react-virtuoso LogViewer companion (#8862)\n* bf2189614 chore: revert multirm refactors (#8962)\n* 4309f7f85 feat: Display resource managers information (#8951)\n* 4d538aec4 test: remove last quarantined test (#8922)\n* 68017dd0d ci: update performance test script for breaking Determined change (#8961)\n* b2b85d718 chore: [RM-68] improve readability for unit test (#8950)\n* e1ca24279 feat: Connect ProjectStore with streaming updates (#8834)\n* 191a14482 fix: don't access agentState when it may be nil (#8921)\n* dcaa8936b fix: update default aux container limits and instance types (#8959)\n* c7e5d4399 docs: fix pre_publish check (#8957)\n* 6ecd81e77 chore: update AMIs - Nvidia minor version bump (#8945)\n* e108ed708 chore: set CGO_ENABLED=0 (#8941)\n* 54aa739dd chore: fix multirm unit test flake (#8949)\n* e8b016541 chore: add resource manager name\u002Fmetadata to resourcepoolv1 proto (#8948)\n* a5b425a82 test: Add e2e test for streaming updates python client (#8901)\n* 77d1ede33 fix: `no data plot` in chart with data (#8935)\n* f416354a4 test: refactor usage within test_local (#8913)\n* c3012ff92 chore: add multirm module to ResourceManager (#8857)\n* d507edd5c test: CLI workflows in CI use new Python images (#8943)\n* fa856ab9d chore: remove support for Python 3.7; prefer 3.8 (#7329)\n* c404c8e24 fix: [RM-6] remove global max-slots-per-pod default when multiple RMs… (#8938)\n* 0d61d1574 build: bump up ci setup_remote_docker version (#8942)\n* 60436eac6 chore: pin pandas and ray versions for ray tests (#8932)\n* f997cd890 fix: malformed config with gcp up with --initial-user-password (#8936)\n* 967e41ff6 build: bump ci cpu image to latest ubuntu 2004 (#8940)\n* 592a566ba feat: streaming updates python client [MD-246] (#8778)\n* 2dfc4f208 chore: remove unused constant (#8934)\n* b99ad9ff6 fix: det deploy gcp down shouldn't check quotas (#8931)\n* 21fb6d1a5 fix: `det dev curl` support for URLs with curly brackets. (#8930)\n* 225dba365 fix: specify go1.22.0 (#8929)\n* 63adae5e9 fix: cli fails when listing providers [DET-10127] (#8903)\n* 9e8cd68ce fix: slurm launcher authenticates preemption notification (#8928)\n* acded3267 tc: Add release note 8851 (#8864)\n* dffda2770 chore: cover and bunify project functions in postgres_experiments.go (#8912)\n* beac3480b fix: SSO button link target (#8925)\n* 5367f4f14 chore: add codeowners for resource-mgmt team files (#8879)\n* feb73de04 tc: Remove broken link (#8924)\n* cd88bb517 chore: revert pod spec and test changes (#8920)\n* 1dfd6d958 chore: bump up ebs size to 400gb for genai deployments\n* 3a3b668de fix: canonicalize master urls shim code (#8919)\n* ef49195aa test: fix failing Go TestResourceCreationFailed test (#8918)\n* 94c7bfe7d chore: minor tweaks as modev takes over streaming updates (#8909)\n* 392f0542a ci: fix failing nightlies after auth PR (#8904)\n* ad7d26019 chore: fix mp.pool test_streaming_metrics_api (#8917)\n* 8af414892 test:  upload test results to datadog (#8910)\n* aab9b42ae test: remove redundant (and brittle) assertion (#8894)\n* 59385a076 feat: log podspec [DET-9861] (#8899)\n* 6857ecf20 chore: refactor ResourceManager interface for multirm (#8847)\n* 9f4060311 test: skip tests that need to get scheduler type (#8911)\n* 4c506011e chore: upgrade Go from 1.21 -> 1.22 (#8914)\n\n","2024-03-18T18:13:54",{"id":260,"version":261,"summary_zh":262,"released_at":263},247793,"0.29.0","## Release Notes\n[0.29.0](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fblob\u002F0.29.0\u002Fdocs\u002Frelease-notes.rst)\n\n## Changelog\n* 50795701a chore: bump version: 0.29.0-rc4 -> 0.29.0\n* 8fa5b5ab4 docs: add release notes for 0.29.0 (#8955)\n* fffde7fa2 chore: bump version: 0.29.0-rc3 -> 0.29.0-rc4\n* f939a0fe5 fix: `no data plot` in chart with data (#8935)\n* 5a74e3723 build: bump ci cpu image to latest ubuntu 2004 (#8940)\n* ad8475987 build: bump up ci setup_remote_docker version (#8942)\n* f0d9768c4 fix: malformed config with gcp up with --initial-user-password (#8936)\n* 2a61ab3fb chore: bump version: 0.29.0-rc2 -> 0.29.0-rc3\n* 435e90a06 chore: fix mp.pool test_streaming_metrics_api (#8917)\n* 18e2ea44f chore: bump version: 0.29.0-rc1 -> 0.29.0-rc2\n* 799373f10 fix: slurm launcher authenticates preemption notification (#8928)\n* 641174cc0 tc: Add release note 8851 (#8864)\n* f275252cc chore: bump up ebs size to 400gb for genai deployments\n* 8c855b72c fix: SSO button link target (#8925)\n* 8d4acd564 tc: Remove broken link (#8924)\n* 06875df29 fix: canonicalize master urls shim code (#8919)\n* e5ae86567 chore: bump version: 0.29.0-rc0 -> 0.29.0-rc1\n* b847edef6 chore: bump version: 0.29.0-dev0 -> 0.29.0-rc0\n* 28c385c3b chore: lock published urls to preserve redirects\n* cbfd3c299 chore: lock api state for backward compatibility check\n* b30f609b2 chore: bump version: 0.28.2-dev0 -> 0.29.0-dev0\n* ad94c1778 fix: return error from websocket handler if socket id is taken (#8877)\n* 46183898f style: update genai logo on sidebar (#8907)\n* 8f82087e8 test: fix tensorboard reattach k8s flake [RM-39] (#8906)\n* d24b19ab4 test: unquarantine deploy-local tests (#8896)\n* 7c6bec9d5 chore: refactor proto, schema, and jobservice for multiRM (#8875)\n* ca96da19c fix: Genai helm service fix (#8885)\n* a89e51e64 fix: trial comparison text overflow bug fix (#8869)\n* 9817a4dd4 chore: add trigger to abort checkpoint deletion (#8878)\n* 2689b0b88 chore: delete unused functions [RM-41] (#8888)\n* 9a6afd263 docs: Organize docs (#8898)\n* a8ac65767 chore: small build system fixes (#8900)\n* fa98bf354 fix: add missing ci context to preview cluster\n* b15d50863 fix: add deploy last main missing ci context (#8892)\n* b47b4772a chore: cleanup stray comments (#8889)\n* ae082656d feat: force default user passwords for all det deploy and CI clusters [RM-28] (#8851)\n* be1ab8519 fix: unnecessary group related api calls during the initial group page loading (#8882)\n* f37bc3ed8 fix: move e2e_tests changes for slurm test from EE to OSS (#8887)\n* 93ced86f8 fix: add missing check for external sessions on exp launch (#8859)\n* 944732a57 ci: more e2e test fixes (#8881)\n* ab9505cb2 ci: fix e2e tests in ee (#8880)\n* 0bc3106b3 docs: Add llm blog link to home page (#8874)\n* c02932769 docs: add link checker utility (#8738)\n* e1da47157 chore: api's default retry now session's default retry (#8872)\n* 7bb9dbcc8 chore: master config updates for multirm [RM-3, RM-4, RM-5, RM-7, RM-29]  (#8831)\n* f101f3d9c chore: add allocation info for cluster ui [DET-10018] (#8616) (#8876)\n* 72d54bea5 chore: canonicalize master urls everywhere [MLG-878] (#8670)\n* e3709bd7d chore: document internal api errors (#8865)\n* 27a279e62 fix: e2e CPU tests have wrong maxSlotsPerPod number (#8870)\n* 03b9b3065 chore: bunify postgres_jobs.go (#8858)\n* e9ac112d7 build(deps): bump peter-evans\u002Fcreate-or-update-comment from 3 to 4 (#8760)\n* dc3e41e54 Fix broken links (#8825)\n* bccdf0c44 fix: stop allowing multi-container allocations to launch in single agent config (#8833)\n* a1214d7b8 chore: add allocation info for cluster ui [DET-10018] (#8616)\n* 76ec233d7 chore: refactor a bunch of auth-related python (#8347)\n* 66b1e6cce chore: bump version: 0.28.1-dev -> 0.28.2-dev0\n* f250ad911 chore: add docs dropdown link for new version\n* 9d44ca1da docs: add release notes for 0.28.1 (#8861)\n* ac8c44094 fix: allow experiments to configure k8s sidecars (#8854)\n* d07ec4091 ci: fix broken ci due to queue version change (#8853)\n* c656aac75 chore: use npm build for hew (#8845)\n* 6b637506c feat: add a master API to fetch a trial by external id. (#8730)\n* e78a4c02a fix: correctly source bucket region when using minio (#8850)\n* dba5f0ff7 fix: replace `react-window` with `react-virtuoso` in transfer component (#8800)\n* 2a183dadd ci: fix performance feature branch using wrong db (#8835)\n* 47061fa42 fix: revert config work from #8765 and #8789 due to feature regressions (#8849)\n* a5f38cbc7 chore: remove GetAllocationSummary from RM interface (#8846)\n* de28a571b chore: cover postgres_jobs.go (#8841)\n* ba8250a8d chore: update backend coverage target (#8798)\n* 556639d21 fix: show error message from backend API for workspace deletion (#8848)\n* 08dfa43af fix: job queue test failures (#8843)\n* 876f9c341 chore: configure agent log level through config file (#8819)\n* ba0337527 chore: move project id onto runs (#8794)\n\n","2024-03-05T18:48:56",{"id":265,"version":266,"summary_zh":267,"released_at":268},247794,"0.28.1","## Release Notes\n[0.28.1](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fblob\u002F0.28.1\u002Fdocs\u002Frelease-notes.rst)\n\n## Changelog\n* f6cb624b4 chore: bump version: 0.28.1-rc3 -> 0.28.1\n* baaa3bd46 docs: add release notes for 0.28.1 (#8861)\n* fbf9df4ca chore: bump version: 0.28.1-rc2 -> 0.28.1-rc3\n* a965f1506 ci: fix broken ci due to queue version change (#8853)\n* d91e8b00f chore: bump version: 0.28.1-rc1 -> 0.28.1-rc2\n* 3129d339b fix: revert config work from #8765 and #8789 due to feature regressions (#8849)\n* 1888b90c8 chore: bump version: 0.28.1-rc0 -> 0.28.1-rc1\n* c443073cf fix: show error message from backend API for workspace deletion (#8848)\n* 1fc1496a7 fix: job queue test failures (#8843)\n* a74685f0a chore: bump version: 0.28.1-dev -> 0.28.1-rc0\n* 5b2e32d90 chore: cleanup the last traces of experiment git fields. [MD-258] (#8830)\n* 92a380f39 feat: Generic task restore (#8802)\n* b0fa7dc82 feat: generic tasks: support startup hooks (#8840)\n* ca80022b6 chore: bunify postgres_checkpoints and add tests (#8783)\n* a4dbc0350 chore: fix error on terminating experiments on restart (#8837)\n* aa98d8262 chore: agent state wasn't getting deleted and logged error (#8838)\n* bb469fae0 fix: update hew with bugfixes (#8839)\n* 393cfde76 Fix broken ref (#8836)\n* 7a1386331 perf: improve GetExperiments + SearchExperiments counting (#8801)\n* d8d996568 chore: remove unused SetAllocationName (#8829)\n* 1946d9a40 docs: Update slurm install (#8832)\n* 1fd21e7fb fix: Fix small typo in Webhook documentation (#8820)\n* e341e2785 feat: Generic Tasks (#8724)\n* fff85e361 fix: handle helm templating in older go template versions (#8828)\n* f300d9797 chore: hide genai helm values config and fix var name (#8821)\n* 6206bde27 feat: add streaming updates core functionality and project streaming (#8669)\n* ed6112186 fix: stop truncating log timestamps to avoid missing logs [WEB-1791] (#8815)\n* 43d3f2119 fix: check for models before deleting workspace (#8804)\n* bb59fa24e ci: wait longer for performance test db to startup (#8796)\n* cfffe96c9 docs: Remove legacy pages (#8818)\n* 1c3f3c485 fix: mitigate many unnecessary api calls in user management table (#8816)\n* 4612c41ca fix: agent config precedence (#8656)\n* 762fcefb4 feat: Deploy GenAI in Helm (#8727)\n* 8e067d9a0 fix: remove possible hang from ship_logs.py [MLG-1565] (#8803)\n* 1daf9d3cc docs: remove duplicated note (#8813)\n* 56e70009c fix: remove extra quotes around IdentifyTask (#8792)\n* 3805ebd89 chore: add testing for k8s informer panic (#8810)\n* a35696d57 refactor: condense trial update functions (#8808)\n* 45c578b9d chore: bump version: 0.28.0-dev0 -> 0.28.1-dev\n* 652062900 chore: add docs dropdown link for new version\n* ed2136d76 docs: add release notes for 0.28.0 (#8807)\n* 825856580 chore: bump version: 0.27.2-dev0 -> 0.28.0-dev\n* c5afb6c42 fix: fetch experiment in case config data is not contained (#8789)\n* 4e17ef74d chore: differentiate between programmatic and web page requests (#8795)\n* a1a6e2074 chore: add ee helm chart changes to oss (#8799)\n* 65c811c09 docs: Add mention of RPMs to on-prem _index.rst (#8773)\n* ad765d4e6 docs: adds\u002Fcorrects EE changes, merges to OSS (#8788)\n* f1a45aee2 perf: update proto_checkpoint_view to use index (#8793)\n* abd590dc3 Revert \"docs: Update oidc and saml docs (#8777)\" (#8791)\n* bb88b01b7 fix: improve trial log request cancelling (#8787)\n* 17f305f54 ci: make perf tests only alert on failure (#8790)\n* 422f5aa44 perf: avoid loading model def in experiment model (#8742)\n* 769845293 perf: improve GetExperiments showTrialData performance (#8753)\n* e801cfe27 perf: add index to checkpoints_v2 id (#8758)\n* 71db4e17b perf: add indexes to tasks and allocations (#8757)\n* e0e6cf0e9 perf: improve get_workspaces query (#8751)\n* ef656bcc2 perf: improve resource agg performance (#8735)\n* e873381d2 fix: retry watcher failure causes infinite loop (#8786)\n* ba2f19053 fix: replace experiment config (#8765)\n* 85d105306 chore: rename postgres_command_intg_test.go (#8785)\n* 40a70cf86 test: performance test CI work (#8761)\n* 36a2e2906 chore: bunify db\u002Fpostgres_tasks.go  (#8764)\n* 07494cf7a fix: update hew to a version without broken documentcard prompts (#8782)\n* 950205992 feat: GCS client should retry on `TooManyRequests`. (#8780)\n* ec850aecd test: add intg tests for db\u002Fpostgres_tasks.go (#8750)\n* 9ec2f7d2f chore: update gke version to comply with latest release for e2e tests (#8781)\n* cefa24213 chore: persist checkpoint storage backend ID (#8690)\n* 905e449dd chore: migrate db schema trials to runs (#8723)\n* dfbb926c0 chore: clean up leftover debug print statements (#8755)\n\n","2024-02-20T22:58:40",{"id":270,"version":271,"summary_zh":272,"released_at":273},247795,"0.28.0","## Release Notes\n[0.28.0](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fblob\u002F0.28.0\u002Fdocs\u002Frelease-notes.rst)\n\n## Changelog\n* 7f9b08272 chore: bump version: 0.28.0-rc4 -> 0.28.0\n* ed1b7f032 docs: add release notes for 0.28.0 (#8807)\n* c4b6f5724 chore: bump version: 0.28.0-rc3 -> 0.28.0-rc4\n* 27ce0a28b chore: add ee helm chart changes to oss (#8799)\n* 959a0968a chore: add ee helm chart changes to oss (#8799)\n* f513174e7 chore: bump version: 0.28.0-rc2 -> 0.28.0-rc3\n* 083c31438 chore: bump version: 0.28.0-rc1 -> 0.28.0-rc2\n* e272bf08b chore: bump version: 0.28.0-rc0 -> 0.28.0-rc1\n* 6080d3918 chore: bump version: 0.27.2-rc4 -> 0.28.0-rc0\n* 3cbce1deb chore: bump version: 0.27.2-rc3 -> 0.27.2-rc4\n* 89df98b8f docs: adds\u002Fcorrects EE changes, merges to OSS (#8788)\n* 2e27c7140 chore: bump version: 0.27.2-rc2 -> 0.27.2-rc3\n* 1abe34f7f chore: bump version: 0.27.2-rc1 -> 0.27.2-rc2\n* e23e1621f fix: improve trial log request cancelling (#8787)\n* 55b5bd42b chore: bump version: 0.27.2-rc0 -> 0.27.2-rc1\n* 5edfd81e5 fix: retry watcher failure causes infinite loop (#8786)\n* 6a21d4440 fix: update hew to a version without broken documentcard prompts (#8782)\n* 74e341dc5 chore: bump version: 0.27.2-dev0 -> 0.27.2-rc0\n* ea9e90372 chore: lock published urls to preserve redirects\n* 0321e1f80 chore: lock api state for backward compatibility check\n* 3783f2b46 docs: Update oidc and saml docs (#8777)\n* 141afa4c9 docs: update dependency version in contributing readme (#8776)\n* 994527f91 fix: Text filter on ProjectMoveModal (#8775)\n* aa65c0773 chore: use vite-plugin-svg-to-jsx package (#8772)\n* 98c61f3d6 test: do not import model_hub test requirements (#8771)\n* 1e2da1054 ci: retry git fetch for early stopping checks (#6318)\n* c73712b57 docs: Replace basic quickstart (#8770)\n* fda515d33 fix: python requirements for pytest and moto (#8769)\n* 78929c09c fix: Filter value resets when switching column types [WEB-1949] (#8731)\n* 7ddf965af docs: Fix minor issues (#8768)\n* 31f6f991e fix: add default transport to proxy connection (#8767)\n* 149b7faca build(deps): bump slackapi\u002Fslack-github-action from 1.24.0 to 1.25.0 (#8766)\n* 719169a31 docs: Fix dropdown url (#8763)\n* 56406a215 Update helm chart config ref (#8762)\n* 5973f8ec6 chore: bump version: 0.27.1-dev0 -> 0.27.2-dev0\n* 260c2bcfd chore: add docs dropdown link for new version\n* 4b4d14adb docs: add release notes for 0.27.1 (#8746)\n* 7841d9e6e feat: the new quick start guide link (#8759)\n* 995311aa4 feat: expconf flag to force scheduling on a single node\u002Fcontainer\u002Fpod (#8743)\n* 64d588f3a refactor: use hew Tree and Divider components [WEB-1920] (#8736)\n* f771acb4f fix: cease many model fetch api calls in checkpoint tab (#8749)\n* 96b9064c6 docs: Add qs for webui users (#8754)\n* d68ffaaca docs: API deprecate returning config for bulk endpoints (#8732)\n* f21a51686 tests: cover queries inside internal\u002Fusers\u002Fpostgres_users.go (#8729)\n* 90a57cb02 fix: Experiment table, right-click context menu [WEB-1942] (#8756)\n* 2ffc18f97 chore: import missing EE helm chart change [ci skip] (#8747)\n* 87b6cf3df fix: use the new genai docker repo (#8745)\n* 7c3650f8d chore: `make devcluster` to rebuild bindings before harness and webui. (#8748)\n* 7f3ddfb3e feat: Add a modal to enable\u002Fdisable Agents [WEB-1718] (#8721)\n* bd0a9ea6e fix: pagination fix in model detail page (#8744)\n* 43c074ec5 feat: helm option to mount shared_fs checkpoints to master (#8741)\n* 9f06d3578 fix: use selected checkpoints when registering (#8739)\n* 6db8c067a test: cover agent_state.go SQL queries (#8740)\n* eb48302f1 test: cover db.GroupCheckpointUUIDsByExperimentID (#8508)\n* d66140442 fix: compress data from API for the page load performance improvement (#8720)\n* b71da7af2 fix: batch metric writes to TensorBoard [MLG-990] (#8688)\n* bd78ec1b6 feat: Preserve 'redirect' query during logout [GAS-489] (#8728)\n* 62941a220 refactor: remove antd App component [WEB-1922] (#8713)\n* bf5b1d1b3 chore: fix unused-imports warning in protos build. (#8726)\n* 0bda0d94f fix: use hew Alert [WEB-1918] (#8711)\n* 1c21f6afa chore: Move from internal glide-table-grid to v6.0.0 [WEB-1945] (#8725)\n* f32e015c4 fix: local checkpoint download path fix (#8722)\n* 190af1d10 docs: [FE-270] add PBS known issue - Cluster tab does not display GPU information (#8719)\n* 6d744f783 feat: content-length for tar checkpoint downloads (#8684)\n* 11e3ba922 chore: upgrade `vitest@1.2.1` (#8718)\n* 92fe3a695 docs: [FE-269] Add documentation detailing configuration steps to set the values for ngpus. (#8714)\n* b69a49c79 chore: update github path in docker docs (#8687)\n* faea5534a chore: codecov reports to match go coverage reports (#8696)\n* 0782c357b chore: standardize oidc\u002Fsaml group & display attribute names in helm config (#8689)\n* acca43417 chore: update oss\u002Fee oidc & saml helm config (#8680)\n* 7188b69ab fix: use Hew dropdown on FilterGroup [WEB-1938] (#8715)\n* a410c4513 chore: Upgrade to vite 5 (#8676)\n* dbeb4581e fix: support `CommandState` for experiment icon (#8709)\n* 83fe47447 docs: fix refe","2024-02-06T20:56:50",{"id":275,"version":276,"summary_zh":277,"released_at":278},247796,"0.27.1","## Release Notes\n[0.27.1](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fblob\u002F0.27.1\u002Fdocs\u002Frelease-notes.rst)\n\n## Changelog\n* e05d57dfd chore: bump version: 0.27.1-rc1 -> 0.27.1\n* 94316ae06 docs: add release notes for 0.27.1 (#8746)\n* b8a772a4c feat: helm option to mount shared_fs checkpoints to master (#8741)\n* 3abd7c7c6 fix: use selected checkpoints when registering (#8739)\n* c479a28dc chore: bump version: 0.27.1-rc0 -> 0.27.1-rc1\n* 6e58dca38 fix: playwright fix (#8699)\n* 204c32efd fix: use Hew dropdown on FilterGroup [WEB-1938] (#8715)\n* f2baf3965 fix: support `CommandState` for experiment icon (#8709)\n* 7a9b0fd6d fix: Update hew for chart fix, avoid error from Typography.Label (#8712)\n* 4085b420f fix: fix CreateExperiment for Remote Users (#8700)\n* 23574cff2 chore: bump version: 0.27.1-dev0 -> 0.27.1-rc0\n* e2cf9809d chore: bumpenvs 0.27.1 (#8701)\n* 4fffb09dc docs: Update references to RHEL and CentOS to Enterprise Linux (#8660)\n* 6ccc7dca8 docs: Fix broken ext link (#8697)\n* 06d766903 docs: fix a few hyperlinks in the Python SDK reference. (#8693)\n* e0558f9c7 fix: Workspace icons display redundant tooltips [WEB-1912] (#8677)\n* 032966753 fix: docs version switcher [MLG-1524] (#8692)\n* 130ebb7cb fix: Trial APIs `--local` should respect `.detignore`. [MLG-1352] (#8683)\n* 3d69af246 test: make compute_stats have a min tolerance of 3 seconds (#8691)\n* 4a30ce80b chore: improve error message for continuing a completed hp search [DET-10041] (#8636)\n* 11a0f2768 fix: add zmq heartbeat in DistributedContext [MLG-1133] (#8681)\n* 10aaf0251 chore: update helm image path to use main branch (#8686)\n* 8671fbd77 feat: set jupyter notebook file browser root to `\u002F`. (#8678)\n* 2fdda3760 fix: update hf trainer api example (#8685)\n* 4118806e5 fix: HP ScatterPlots scrollbar (#8682)\n* 248e88c59 fix: update helm chart for new logo (#8675)\n* 4e8d9a94a feat: Truncate and pad cell values in glide table [WEB-1778] (#8665)\n* 38e68253e fix: Remove out of place spinner at dashboard page (#8668)\n* df67efee8 fix: use FixedSizeList for column picker to fix jitter (#8628)\n* 127cf98ed fix: break out stopping... states from stopped states in webui (#8672)\n* 766577d96 feat: add health check for genai deployments (#8613)\n* 0d938552b feat: `det (notebook|tensorboard|shell|command) ls --sort-by ...` [DET-6126] (#8649)\n* 39ad129d2 fix: wait on algolia search upload (#8661)\n* f5bebc333 chore: delete unused project model (#8673)\n* c99bfc3a8 fix: Helm master-service clusterip mngt. openshift (#8546)\n* 4354c8edb chore: reinstate original DuplicateError to postgres_users.go (#8657)\n* 67a2c4087 chore: bump version: 0.27.0-dev0 -> 0.27.1-dev0\n* 8446c43b9 chore: add docs dropdown link for new version\n* 8651f411b docs: add release notes for 0.27.0 (#8671)\n* 25f0ba9ef chore: Add warning to Patch Mast config to specify changes are ephemeral (#8577)\n* fd4c1a0c2 fix: don't drop active_user on expired tokens [MLG-1494] (#8653)\n* 9dc4fc83f feat: add `det e unpause` alias. (#8562)\n* 7d9f29532 style: reformat `det (n|t|s|c)` arguments code. (#8652)\n* c1dbed911 chore: Remove antd usage from determined [WEB-1723] (#8605)\n* 0c9e19f0e feat: conditionally add genai to the sidebar (#8496)\n* 4f4ef5a34 fix: stop using distutils (for python 3.12) [MLG-1519] (#8667)\n* 1e8baee4a fix: Add theme to TableFilterDropdown [WEB-1952] (#8663)\n* 24cfbb83b docs: Add section on viewing topology (#8638)\n* ef26ad2ed fix: Revert \"chore: Job\u002Ftask displays Running instead of Scheduled (#8335)\" (#8654)\n* 926a6de51 chore: delete unused experiment fields in db (#8639)\n* 33dcfbd7f fix: replace View Logs link with useNavigate (#8655)\n* 87f713fc8 fix: Check if loading before saying no workspaces \u002F projects [WEB-1904] (#8627)\n* ae072f70e docs: Reorder the tutorials for visibility (#8650)\n* 5774f5cc3 chore: improve string_to_bool help text of cli [MLG-1208] (#8651)\n* ab0834542 chore: move postgres_command.go to command module (#8648)\n* c1a746116 chore: add Go coverage ratchet to unit and integration tests (#8602)\n* c0614a8cb fix(cli\u002Fdeploy\u002Fsimple-rds): set template default snapshot (#8645)\n* 2403cd5fc fix: Sort order in users.active column (#8646)\n* b60f8de54 fix: Trial metric chart series always have unique colors (#8626)\n* d1b585c9d test: fix `test_tf_keras_parallel`. (#8641)\n* 5f19a534f feat(cli\u002Fdeploy): add db snapshot flag for simple-rds [INFENG-269] (#8443)\n* 2e6a634c4 chore: print TB status message to logs [MLG-1119] (#8642)\n* c8d50a3e6 fix: ResourcepoolDetails polls for job stats [WEB-1914] (#8635)\n* 2604235fc chore: Update hew to 0.6.22 (#8634)\n* 34891a11e docs: Add oidc note about uniqueness (#8637)\n* 001d82708 chore: extend CLI NTSC timeout [MLG-870] (#8632)\n* 86aab147b docs: Document scim display name attribute (#8606)\n* 134c2e149 fix: reduce tfkeras iris const batch size (#8633)\n* a5e4b706f docs: Mention examples in the git readme (#8623)\n\n","2024-01-24T23:32:42",{"id":280,"version":281,"summary_zh":282,"released_at":283},247797,"0.27.0","## Release Notes\n[0.27.0](https:\u002F\u002Fgithub.com\u002Fdetermined-ai\u002Fdetermined\u002Fblob\u002F0.27.0\u002Fdocs\u002Frelease-notes.rst)\n\n## Changelog\n* eb0eae8af chore: bump version: 0.27.0-rc4 -> 0.27.0\n* b51bfc550 docs: add release notes for 0.27.0 (#8671)\n* 15783b5fb Revert \"adding 0.27 release notes\"\n* d765c6e5f revert doc changes\n* 13727ad0d Revert doc changes:\n* e96297717 Revert \"revert doc changes\"\n* a5e284993 revert doc changes\n* e8ee69db2 adding 0.27 release notes\n* bc7c75bcc chore: bump version: 0.27.0-rc3 -> 0.27.0-rc4\n* 69b93a2a5 fix: Add theme to TableFilterDropdown [WEB-1952] (#8663)\n* bb12b9395 chore: bump version: 0.27.0-rc2 -> 0.27.0-rc3\n* f3234099d fix: ResourcepoolDetails polls for job stats [WEB-1914] (#8635)\n* 1785b44c7 chore: bump version: 0.27.0-rc1 -> 0.27.0-rc2\n* 35e4b4949 chore: Update hew to 0.6.22 (#8634)\n* b899482b7 chore: bump version: 0.27.0-rc0 -> 0.27.0-rc1\n* 633d455d2 chore: bump version: 0.27.0-dev0 -> 0.27.0-rc0\n* 6e6a23e35 chore: add docs dropdown link for new version\n* 0bc1644b4 chore: bump version: 0.26.8-dev0 -> 0.27.0-dev0\n* f0e4c0c7e chore: added release note for breaking CLI experiment creation change (#8630)\n* eea7ca566 fix: remove UNSPECIFIED values from end-user enums (#8367)\n* 9de3e4ba5 fix: os.uname() not found on windows (#8629)\n* 78068eef8 chore: cli describe workspaces no longer also lists projects (#8609)\n* 79df355a2 chore: delete unused trials_augmented_view (#8588)\n* edd71b315 chore: update hew to 0.6.20 (#8625)\n* 1818cb481 ci: fix test split for skipped tests (#8624)\n* fa8bf666b chore: fix a typo in RRI (#8612)\n* fac9fa0b4 fix: update experiment hp search param quoting (#8620)\n* 601a98f35 docs: Fix mmdetection readme link (#8607)\n* b50a81042 build(deps): bump golang.org\u002Fx\u002Fcrypto from 0.14.0 to 0.17.0 (#8604)\n* 8d175e69a chore(deps): bump flake8-commas from 2.0.0 to 2.1.0 (#4338)\n* 9b88de4ad build(deps): bump actions\u002Fsetup-go from 4 to 5 (#8556)\n* 590c0b919 build(deps): bump actions\u002Fsetup-python from 4 to 5 (#8555)\n* 27e3c1d7e feat: Replace useModal with updated Modal component [WEB-1828] [WEB-1891] (#8578)\n* 63e44b094 Add network connectivity diagram (#8446)\n* 83f564266 chore: Model Version pages to Hew UI [WEB-1814] (#8603)\n* b462a5481 chore: bump version: 0.26.7-dev0 -> 0.26.8-dev0\n* 14ba6bd69 docs: add release notes for 0.26.7 (#8601)\n* c2e7d1db4 build(deps): bump github\u002Fcodeql-action from 2 to 3 (#8594)\n* 490f537c3 chore(deps): bump actions\u002Fdownload-artifact from 3 to 4 (#8598)\n* ded49abf0 chore(deps): bump actions\u002Fupload-artifact from 3 to 4 (#8599)\n* 7ddcb92f7 fix: Show experiment loading state before first loading (#8583)\n* 2c29f998a chore: Remove attrdict from model hub (#8554)\n* 7d0afb7ba chore: Upgrade hew to 0.6.19 (#8597)\n* 0f5a16d9f chore: Change CLI and constants from lore to genai (#8593)\n* ac9ee99fd fix: prevent theme settings conflict in multi-client\u002Fwindow scenario (#8596)\n* 2cd8b3e26 refactor: Replace antd imports with UI kit usage [WEB-1723] (#8582)\n* 576ffb56d fix: antd.Typography with expansion tooltips moved to hew [WEB-1862] (#8513)\n* e37ca8290 chore: drop raw_checkpoints table (#8584)\n* 37199680a fix: Change slot number back to 2 in Keras example (#8595)\n* 22a17d4f9 chore(deps): bump github.com\u002Fdocker\u002Fdocker from 20.10.24+incompatible to 24.0.7+incompatible (#8316)\n* 909c0d5e3 fix: allow for `model_dir` to be None when submitting experiments (#8422)\n* 79321f7ce chore: encode path parameters in generated bindings [MLG-781] (#8543)\n* ce9fe4344 chore: update byExternalToken signature (#8587)\n* dccc52691 test: skip more `mmdetection` tests [DET-10036] (#8592)\n* 6fe2c916e feat: add local checkpoint storage download through server (#8505)\n* e1259556f chore: fix api state (#8589)\n\n","2024-01-10T15:22:09"]