[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-microsoft--mattersim":3,"tool-microsoft--mattersim":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",155373,2,"2026-04-14T11:34:08",[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":66,"readme_en":67,"readme_zh":68,"quickstart_zh":69,"use_case_zh":70,"hero_image_url":71,"owner_login":72,"owner_name":73,"owner_avatar_url":74,"owner_bio":75,"owner_company":76,"owner_location":76,"owner_email":77,"owner_twitter":78,"owner_website":79,"owner_url":80,"languages":81,"stars":94,"forks":95,"last_commit_at":96,"license":97,"difficulty_score":32,"env_os":98,"env_gpu":99,"env_ram":100,"env_deps":101,"category_tags":110,"github_topics":112,"view_count":32,"oss_zip_url":76,"oss_zip_packed_at":76,"status":17,"created_at":121,"updated_at":122,"faqs":123,"releases":152},7449,"microsoft\u002Fmattersim","mattersim","MatterSim: A deep learning atomistic model across elements, temperatures and pressures.","MatterSim 是一款由微软推出的深度学习原子模型，旨在跨越不同化学元素、温度及压力条件，精准模拟材料的微观行为。它主要解决了传统材料模拟方法在计算成本高、适用范围窄以及难以兼顾多物理场条件时的痛点，为研究人员提供了一种高效且通用的替代方案。\n\n这款工具特别适合材料科学家、计算化学研究者以及相关领域的开发者使用。通过预训练的 M3GNet 架构模型，MatterSim 能够快速预测材料的能量、原子受力及应力张量等关键物理性质，无需从头进行昂贵的量子力学计算。其独特亮点在于提供了两个版本的预训练模型：轻量版注重推理速度，适合快速筛选；大参数版则追求更高的模拟精度。此外，MatterSim 支持灵活的部署方式，既可通过 PyPI 一键安装，也支持源码编译，并兼容主流深度学习框架。无论是探索新型合金、电池材料，还是研究极端环境下的物质状态，MatterSim 都能成为加速科学发现的得力助手。","\u003Ch1>\n\u003Cp align=\"center\">\n    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fmicrosoft_mattersim_readme_a250db9ce121.png\" alt=\"MatterSim logo\" width=\"600\"\u002F>\n\u003C\u002Fp>\n\u003C\u002Fh1>\n\n\u003C!-- \u003Ch1 align=\"center\">MatterSim\u003C\u002Fh1> -->\n\n\u003Ch4 align=\"center\">\n\n[![arXiv](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FarXiv-2405.04967-blue?logo=arxiv&logoColor=white.svg)](https:\u002F\u002Farxiv.org\u002Fabs\u002F2405.04967)\n[![Requires Python 3.10+](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FPython-3.10+-blue.svg?logo=python&logoColor=white)](https:\u002F\u002Fpython.org\u002Fdownloads)\n[![PyPI Downloads](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fmicrosoft_mattersim_readme_34c50e933400.png)](https:\u002F\u002Fpepy.tech\u002Fprojects\u002Fmattersim)\n\u003C\u002Fh4>\n\n\nMatterSim is a deep learning atomistic model across elements, temperatures and pressures.\n\n## Documentation\n\nThis README provides a quick start guide. For more comprehensive information, please refer to the [MatterSim documentation](https:\u002F\u002Fmicrosoft.github.io\u002Fmattersim\u002F).\n\n## Installation\n\n### Prerequisite\n* `Python >= 3.10`\n\n\n### Install from PyPI\n> [!TIP]\n> While not mandatory, we recommend creating a clean conda environment before installing MatterSim to avoid potential package conflicts. You can create and activate a conda environment with the following commands:\n>\n> ```bash\n> # create the environment\n> conda create -n mattersim python=3.10\n>\n> # activate the environment\n> conda activate mattersim\n> ```\n>\n\nTo install MatterSim, use the following command. Please note that downloading the dependencies may take some time:\n```bash\npip install mattersim\n```\n\nIn case you want to install the package with the latest version, you can run the following command:\n\n```bash\npip install git+https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim.git\n```\n\n### Install from source code\n1. Download the source code of MatterSim and change to the directory\n\n```bash\ngit clone git@github.com:microsoft\u002Fmattersim.git\ncd mattersim\n```\n\n2. Install MatterSim\n\n> [!WARNING]\n> We strongly recommend that users install MatterSim using [mamba or micromamba](https:\u002F\u002Fmamba.readthedocs.io\u002Fen\u002Flatest\u002Findex.html), because *conda* can be significantly slower when resolving the dependencies in environment.yaml.\n\nTo install the package, run the following command under the root of the folder:\n\n```bash\nmamba env create -f environment.yaml\nmamba activate mattersim\nuv pip install -e .\n```\n\n## Pre-trained Models\n\nWe currently offer two pre-trained **MatterSim-v1** models based on the **M3GNet** architecture in the [pretrained_models](.\u002Fpretrained_models\u002F) folder:\n\n1. **MatterSim-v1.0.0-1M**: A mini version of the model that is faster to run.\n2. **MatterSim-v1.0.0-5M**: A larger version of the model that is more accurate.\n\nThese models have been trained using the data generated through the workflows\nintroduced in the [MatterSim manuscript](https:\u002F\u002Farxiv.org\u002Fabs\u002F2405.04967), which provides an in-depth\nexplanation of the methodologies underlying the MatterSim model.\n\nMore advanced and fully-supported pretrained versions of MatterSim,\nand additional materials capabilities are available in\n**[Azure Quantum Elements](https:\u002F\u002Fquantum.microsoft.com\u002Fen-us\u002Fsolutions\u002Fazure-quantum-elements)**.\n\n## Usage\n\n> [!TIP]\n> **Note for macOS Users:** If you are using macOS with Apple Silicon, please be aware of potential numerical instability with the MPS backend. We recommend using the CPU device for MatterSim on Mac to avoid these issues.\n\n### A minimal test\n```python\nimport torch\nfrom loguru import logger\nfrom ase.build import bulk\nfrom ase.units import GPa\nfrom mattersim.forcefield import MatterSimCalculator\n\ndevice = \"cuda\" if torch.cuda.is_available() else \"cpu\"\nlogger.info(f\"Running MatterSim on {device}\")\n\nsi = bulk(\"Si\", \"diamond\", a=5.43)\nsi.calc = MatterSimCalculator(device=device)\nlogger.info(f\"Energy (eV)                 = {si.get_potential_energy()}\")\nlogger.info(f\"Energy per atom (eV\u002Fatom)   = {si.get_potential_energy()\u002Flen(si)}\")\nlogger.info(f\"Forces of first atom (eV\u002FA) = {si.get_forces()[0]}\")\nlogger.info(f\"Stress[0][0] (eV\u002FA^3)       = {si.get_stress(voigt=False)[0][0]}\")\nlogger.info(f\"Stress[0][0] (GPa)          = {si.get_stress(voigt=False)[0][0] \u002F GPa}\")\n```\n\nIn this release, we provide two checkpoints: `MatterSim-v1.0.0-1M.pth` and `MatterSim-v1.0.0-5M.pth`. By default, the `1M` version is loaded.\nTo switch to the `5M` version, manually set the `load_path` of `MatterSimCalculator` as shown below:\n\n```python\nMatterSimCalculator(load_path=\"MatterSim-v1.0.0-5M.pth\", device=device)\n```\n\n## Finetune\n> [!TIP]\n> MatterSim provides a finetune script to finetune the pre-trained MatterSim model on a custom dataset.\n> Please refer to the [MatterSim documentation](https:\u002F\u002Fmicrosoft.github.io\u002Fmattersim\u002F) for more details.\n\n### A minimal finetune example\n\n```bash\ntorchrun --nproc_per_node=1 src\u002Fmattersim\u002Ftraining\u002Ffinetune_mattersim.py --load_model_path mattersim-v1.0.0-1m --train_data_path tests\u002Fdata\u002Fhigh_level_water.xyz\n```\n\n\n\n\n## Reference\n\nWe kindly request that users of MatterSim version 1.0.0 cite our preprint available on arXiv:\n```\n@article{yang2024mattersim,\n      title={MatterSim: A Deep Learning Atomistic Model Across Elements, Temperatures and Pressures},\n      author={Han Yang and Chenxi Hu and Yichi Zhou and Xixian Liu and Yu Shi and Jielan Li and Guanzhi Li and Zekun Chen and Shuizhou Chen and Claudio Zeni and Matthew Horton and Robert Pinsler and Andrew Fowler and Daniel Zügner and Tian Xie and Jake Smith and Lixin Sun and Qian Wang and Lingyu Kong and Chang Liu and Hongxia Hao and Ziheng Lu},\n      year={2024},\n      eprint={2405.04967},\n      archivePrefix={arXiv},\n      primaryClass={cond-mat.mtrl-sci},\n      url={https:\u002F\u002Farxiv.org\u002Fabs\u002F2405.04967},\n      journal={arXiv preprint arXiv:2405.04967}\n}\n```\n\n> [!IMPORTANT]\n> We kindly ask users to **explicitly** specify the exact model version and checkpoint (e.g., **MatterSim-v1.0.0-1M**) when reporting results in academic papers or technical reports, rather than referring to the model generically as **MatterSim**. Precise versioning is crucial for ensuring reproducibility. For instance, the statement \"_This study was conducted using MatterSim-v1.0.0-1M_\" serves as a good example.\n\n## Limitations\n\n**MatterSim-v1** is designed specifically for atomistic simulations of bulk materials. Applications or interpretations beyond this scope should be approached with caution. For instance, when using the model for simulations involving surfaces, interfaces, or properties influenced by long-range interactions, the results may be qualitatively accurate but are not suitable for quantitative analysis. In such cases, we recommend fine-tuning the model to better align with the specific application.\n\n## Trademarks\n\nThis project may contain trademarks or logos for projects, products, or services.\nAuthorized use of Microsoft trademarks or logos is subject to and must follow [Microsoft's Trademark & Brand Guidelines](https:\u002F\u002Fwww.microsoft.com\u002Fen-us\u002Flegal\u002Fintellectualproperty\u002Ftrademarks\u002Fusage\u002Fgeneral).\nUse of Microsoft trademarks or logos in modified versions of this project must not cause confusion or imply Microsoft sponsorship.\nAny use of third-party trademarks or logos are subject to those third-party's policies.\n\n## Responsible AI Transparency Documentation\n\nThe responsible AI transparency documentation can be found [here](MODEL_CARD.md).\n\n\n## Researcher and Developers\nMatterSim is actively under development, and we welcome community engagement. If you have research interests related to this model, ideas you’d like to contribute, or issues to report, we encourage you to reach out to us at [ai4s-materials@microsoft.com](mailto:ai4s-materials@microsoft.com).\n","\u003Ch1>\n\u003Cp align=\"center\">\n    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fmicrosoft_mattersim_readme_a250db9ce121.png\" alt=\"MatterSim logo\" width=\"600\"\u002F>\n\u003C\u002Fp>\n\u003C\u002Fh1>\n\n\u003C!-- \u003Ch1 align=\"center\">MatterSim\u003C\u002Fh1> -->\n\n\u003Ch4 align=\"center\">\n\n[![arXiv](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FarXiv-2405.04967-blue?logo=arxiv&logoColor=white.svg)](https:\u002F\u002Farxiv.org\u002Fabs\u002F2405.04967)\n[![Requires Python 3.10+](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FPython-3.10+-blue.svg?logo=python&logoColor=white)](https:\u002F\u002Fpython.org\u002Fdownloads)\n[![PyPI Downloads](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fmicrosoft_mattersim_readme_34c50e933400.png)](https:\u002F\u002Fpepy.tech\u002Fprojects\u002Fmattersim)\n\u003C\u002Fh4>\n\n\nMatterSim 是一种跨元素、温度和压力的深度学习原子级模型。\n\n## 文档\n\n本 README 提供快速入门指南。如需更全面的信息，请参阅 [MatterSim 文档](https:\u002F\u002Fmicrosoft.github.io\u002Fmattersim\u002F)。\n\n## 安装\n\n### 先决条件\n* `Python >= 3.10`\n\n\n### 从 PyPI 安装\n> [!TIP]\n> 虽然不是强制要求，但我们建议在安装 MatterSim 之前创建一个干净的 conda 环境，以避免潜在的包冲突。您可以通过以下命令创建并激活 conda 环境：\n>\n> ```bash\n> # 创建环境\n> conda create -n mattersim python=3.10\n>\n> # 激活环境\n> conda activate mattersim\n> ```\n>\n\n要安装 MatterSim，请使用以下命令。请注意，下载依赖项可能需要一些时间：\n```bash\npip install mattersim\n```\n\n如果您希望安装最新版本的软件包，可以运行以下命令：\n\n```bash\npip install git+https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim.git\n```\n\n### 从源代码安装\n1. 下载 MatterSim 的源代码并切换到该目录\n\n```bash\ngit clone git@github.com:microsoft\u002Fmattersim.git\ncd mattersim\n```\n\n2. 安装 MatterSim\n\n> [!WARNING]\n> 我们强烈建议用户使用 [mamba 或 micromamba](https:\u002F\u002Fmamba.readthedocs.io\u002Fen\u002Flatest\u002Findex.html) 来安装 MatterSim，因为在解析 environment.yaml 中的依赖关系时，*conda* 的速度可能会显著变慢。\n\n要安装软件包，请在文件夹根目录下运行以下命令：\n\n```bash\nmamba env create -f environment.yaml\nmamba activate mattersim\nuv pip install -e .\n```\n\n## 预训练模型\n\n我们目前在 [pretrained_models](.\u002Fpretrained_models\u002F) 文件夹中提供了两个基于 **M3GNet** 架构的预训练 **MatterSim-v1** 模型：\n\n1. **MatterSim-v1.0.0-1M**：模型的迷你版本，运行速度更快。\n2. **MatterSim-v1.0.0-5M**：模型的较大版本，精度更高。\n\n这些模型是使用在 [MatterSim 论文](https:\u002F\u002Farxiv.org\u002Fabs\u002F2405.04967) 中介绍的工作流程生成的数据进行训练的，该论文深入解释了 MatterSim 模型背后的方法论。\n\n更多高级且完全支持的 MatterSim 预训练版本以及额外的材料功能，可在 **[Azure Quantum Elements](https:\u002F\u002Fquantum.microsoft.com\u002Fen-us\u002Fsolutions\u002Fazure-quantum-elements)** 中获得。\n\n## 使用\n\n> [!TIP]\n> **macOS 用户注意：** 如果您使用的是搭载 Apple Silicon 的 macOS，请注意 MPS 后端可能存在数值不稳定性。为避免这些问题，我们建议在 Mac 上使用 CPU 设备运行 MatterSim。\n\n### 一个最小测试\n```python\nimport torch\nfrom loguru import logger\nfrom ase.build import bulk\nfrom ase.units import GPa\nfrom mattersim.forcefield import MatterSimCalculator\n\ndevice = \"cuda\" if torch.cuda.is_available() else \"cpu\"\nlogger.info(f\"Running MatterSim on {device}\")\n\nsi = bulk(\"Si\", \"diamond\", a=5.43)\nsi.calc = MatterSimCalculator(device=device)\nlogger.info(f\"Energy (eV)                 = {si.get_potential_energy()}\")\nlogger.info(f\"Energy per atom (eV\u002Fatom)   = {si.get_potential_energy()\u002Flen(si)}\")\nlogger.info(f\"Forces of first atom (eV\u002FA) = {si.get_forces()[0]}\")\nlogger.info(f\"Stress[0][0] (eV\u002FA^3)       = {si.get_stress(voigt=False)[0][0]}\")\nlogger.info(f\"Stress[0][0] (GPa)          = {si.get_stress(voigt=False)[0][0] \u002F GPa}\")\n```\n\n在此版本中，我们提供了两个检查点：`MatterSim-v1.0.0-1M.pth` 和 `MatterSim-v1.0.0-5M.pth`。默认加载的是 `1M` 版本。要切换到 `5M` 版本，可手动设置 `MatterSimCalculator` 的 `load_path`，如下所示：\n\n```python\nMatterSimCalculator(load_path=\"MatterSim-v1.0.0-5M.pth\", device=device)\n```\n\n## 微调\n> [!TIP]\n> MatterSim 提供了一个微调脚本，可用于在自定义数据集上对预训练的 MatterSim 模型进行微调。有关详细信息，请参阅 [MatterSim 文档](https:\u002F\u002Fmicrosoft.github.io\u002Fmattersim\u002F)。\n\n### 一个最小的微调示例\n\n```bash\ntorchrun --nproc_per_node=1 src\u002Fmattersim\u002Ftraining\u002Ffinetune_mattersim.py --load_model_path mattersim-v1.0.0-1m --train_data_path tests\u002Fdata\u002Fhigh_level_water.xyz\n```\n\n\n\n\n## 参考文献\n\n我们恳请使用 MatterSim 1.0.0 版本的用户引用我们在 arXiv 上发布的预印本：\n```\n@article{yang2024mattersim,\n      title={MatterSim: A Deep Learning Atomistic Model Across Elements, Temperatures and Pressures},\n      author={Han Yang and Chenxi Hu and Yichi Zhou and Xixian Liu and Yu Shi and Jielan Li and Guanzhi Li and Zekun Chen and Shuizhou Chen and Claudio Zeni and Matthew Horton and Robert Pinsler and Andrew Fowler and Daniel Zügner and Tian Xie and Jake Smith and Lixin Sun and Qian Wang and Lingyu Kong and Chang Liu and Hongxia Hao and Ziheng Lu},\n      year={2024},\n      eprint={2405.04967},\n      archivePrefix={arXiv},\n      primaryClass={cond-mat.mtrl-sci},\n      url={https:\u002F\u002Farxiv.org\u002Fabs\u002F2405.04967},\n      journal={arXiv preprint arXiv:2405.04967}\n}\n```\n\n> [!IMPORTANT]\n> 我们恳请用户在学术论文或技术报告中报告结果时，**明确**指定确切的模型版本和检查点（例如 **MatterSim-v1.0.0-1M**），而不是笼统地称其为 **MatterSim**。精确的版本控制对于确保结果的可重复性至关重要。例如，“本研究使用 MatterSim-v1.0.0-1M 进行”就是一个很好的例子。\n\n## 局限性\n\n**MatterSim-v1** 专门设计用于块体材料的原子级模拟。超出此范围的应用或解释应谨慎对待。例如，当使用该模型进行涉及表面、界面或受长程相互作用影响的性质的模拟时，结果可能在定性上准确，但不适合定量分析。在这种情况下，我们建议对该模型进行微调，使其更好地适应具体应用。\n\n## 商标\n\n本项目可能包含项目、产品或服务的商标或标识。  \n对微软商标或标识的授权使用须遵守并严格执行[微软商标与品牌指南](https:\u002F\u002Fwww.microsoft.com\u002Fen-us\u002Flegal\u002Fintellectualproperty\u002Ftrademarks\u002Fusage\u002Fgeneral)。  \n在本项目的修改版本中使用微软商标或标识时，不得造成混淆或暗示微软的赞助关系。  \n任何第三方商标或标识的使用均应遵循该第三方的相关政策。\n\n## 负责任的人工智能透明度文档\n\n负责任的人工智能透明度文档可在此处找到：[MODEL_CARD.md]。\n\n\n## 研究人员与开发者\nMatterSim 目前仍在积极开发中，我们欢迎社区参与。如果您对该模型的研究方向感兴趣、有建设性意见或需要报告问题，请随时通过电子邮件 [ai4s-materials@microsoft.com](mailto:ai4s-materials@microsoft.com) 与我们联系。","# MatterSim 快速上手指南\n\nMatterSim 是一个跨元素、温度和压力的深度学习原子模型，适用于体相材料的原子模拟。\n\n## 环境准备\n\n*   **操作系统**：支持 Linux、macOS（Apple Silicon 用户建议使用 CPU 以避免数值不稳定）和 Windows。\n*   **Python 版本**：必须 >= 3.10。\n*   **推荐工具**：建议使用 `conda`、`mamba` 或 `micromamba` 创建独立的虚拟环境，以避免依赖冲突。\n\n## 安装步骤\n\n### 方法一：通过 PyPI 安装（推荐）\n\n首先创建并激活 conda 环境：\n\n```bash\n# 创建环境\nconda create -n mattersim python=3.10\n\n# 激活环境\nconda activate mattersim\n```\n\n安装 MatterSim（下载依赖可能需要一些时间）：\n\n```bash\npip install mattersim\n```\n\n如需安装最新的开发版本：\n\n```bash\npip install git+https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim.git\n```\n\n### 方法二：从源码安装\n\n如果您需要修改源码或使用最新功能，请按以下步骤操作：\n\n1. 克隆仓库并进入目录：\n```bash\ngit clone git@github.com:microsoft\u002Fmattersim.git\ncd mattersim\n```\n\n2. 使用 `mamba` 安装依赖（强烈推荐，因为 `conda` 解析依赖较慢）：\n```bash\nmamba env create -f environment.yaml\nmamba activate mattersim\nuv pip install -e .\n```\n\n## 基本使用\n\n以下是一个最小化的测试示例，用于计算硅（Si）晶体的能量、受力和应力。默认加载 `MatterSim-v1.0.0-1M`（轻量版）模型。\n\n> **注意**：macOS Apple Silicon 用户若遇到数值不稳定问题，请将 `device` 设置为 `\"cpu\"`。\n\n```python\nimport torch\nfrom loguru import logger\nfrom ase.build import bulk\nfrom ase.units import GPa\nfrom mattersim.forcefield import MatterSimCalculator\n\n# 自动检测 CUDA，若无则使用 CPU\ndevice = \"cuda\" if torch.cuda.is_available() else \"cpu\"\nlogger.info(f\"Running MatterSim on {device}\")\n\n# 构建硅晶体结构\nsi = bulk(\"Si\", \"diamond\", a=5.43)\nsi.calc = MatterSimCalculator(device=device)\n\n# 输出计算结果\nlogger.info(f\"Energy (eV)                 = {si.get_potential_energy()}\")\nlogger.info(f\"Energy per atom (eV\u002Fatom)   = {si.get_potential_energy()\u002Flen(si)}\")\nlogger.info(f\"Forces of first atom (eV\u002FA) = {si.get_forces()[0]}\")\nlogger.info(f\"Stress[0][0] (eV\u002FA^3)       = {si.get_stress(voigt=False)[0][0]}\")\nlogger.info(f\"Stress[0][0] (GPa)          = {si.get_stress(voigt=False)[0][0] \u002F GPa}\")\n```\n\n### 切换模型版本\n\n默认加载的是 `1M` 版本。如需使用更精确但较大的 `5M` 版本（`MatterSim-v1.0.0-5M`），请手动指定 `load_path`：\n\n```python\nMatterSimCalculator(load_path=\"MatterSim-v1.0.0-5M.pth\", device=device)\n```","某新能源电池材料研发团队正急需筛选一种能在极端高温高压环境下保持稳定的新型固态电解质，以突破现有电池的安全瓶颈。\n\n### 没有 mattersim 时\n- **计算成本高昂**：依赖传统密度泛函理论（DFT）模拟原子行为，单次针对特定温度和压力的计算需耗时数小时甚至数天，难以覆盖广阔的化学空间。\n- **条件受限严重**：传统方法在模拟极端高温或高压条件时往往收敛困难，甚至无法得到可靠的物理性质数据，导致实验盲测风险大。\n- **元素覆盖狭窄**：现有的快速势函数模型通常只支持少数几种常见元素，一旦涉及多元素掺杂或复杂化合物，就必须重新训练或退回慢速计算。\n- **迭代周期漫长**：从提出假设到验证材料稳定性需要数周时间，严重拖慢了新材料的发现节奏，错失市场先机。\n\n### 使用 mattersim 后\n- **推理速度飞跃**：利用 matterSim 预训练的深度学习模型，团队可在几分钟内完成原本需要数天的原子级模拟，快速评估成千上万种候选材料。\n- **全条件精准预测**：matterSim 原生支持跨元素、宽温度范围和不同压力条件的模拟，轻松输出极端工况下的能量、受力和应力数据，无需额外调整。\n- **通用性强**：凭借跨越多种元素的训练数据，matterSim 直接适用于复杂的多元合金或掺杂体系，无需为每种新组合重新建模。\n- **研发效率倍增**：筛选周期从数周缩短至数小时，研究人员能迅速锁定最有潜力的几种材料进行实验验证，大幅降低试错成本。\n\nmatterSim 通过将高精度的量子力学计算转化为秒级的深度学习推理，让材料科学家能在极端条件下以前所未有的速度探索未知物质世界。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fmicrosoft_mattersim_a250db9c.png","microsoft","Microsoft","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002Fmicrosoft_4900709c.png","Open source projects and samples from Microsoft",null,"opensource@microsoft.com","OpenAtMicrosoft","https:\u002F\u002Fopensource.microsoft.com","https:\u002F\u002Fgithub.com\u002Fmicrosoft",[82,86,90],{"name":83,"color":84,"percentage":85},"Jupyter Notebook","#DA5B0B",57.1,{"name":87,"color":88,"percentage":89},"Python","#3572A5",42.3,{"name":91,"color":92,"percentage":93},"Cython","#fedf5b",0.6,528,78,"2026-04-13T15:57:09","MIT","Linux, macOS","非必需。支持 CUDA (NVIDIA GPU) 或 CPU。macOS Apple Silicon 用户因 MPS 后端数值不稳定问题，建议强制使用 CPU。具体显存大小和 CUDA 版本未在文档中明确说明。","未说明",{"notes":102,"python":103,"dependencies":104},"1. 强烈建议使用 mamba 或 micromamba（而非 conda）来创建环境，因为 conda 解析依赖项速度较慢。2. macOS Apple Silicon 用户请注意：MPS 后端可能存在数值不稳定性，建议在 Mac 上使用 CPU 设备运行。3. 提供两个预训练模型版本：'1M'（小型、快速，默认加载）和 '5M'（大型、更准确）。4. 该模型专为块体材料（bulk materials）的原子模拟设计，用于表面、界面或长程相互作用时需谨慎，可能需微调。","3.10+",[105,106,107,108,109],"torch","ase","loguru","mamba\u002Fmicromamba (推荐用于环境管理)","uv",[14,111],"其他",[113,114,115,116,117,118,119,120],"ai4materials","ai4science","foundation-models","machine-learning-force-field","mlff","computational-materials-science","machine-learning","materials-science","2026-03-27T02:49:30.150509","2026-04-14T20:49:57.357466",[124,129,134,139,144,148],{"id":125,"question_zh":126,"answer_zh":127,"source_url":128},33421,"在使用 MatterSim 计算小分子（非周期性结构）时出现 \"No PBC detected\" 警告或得到 NaN\u002F异常大的结果，该如何处理？","MatterSim 目前对非周期性边界条件（non-PBC）的支持不够稳健。建议的处理方法是：人为构建一个较大的晶胞，将分子放入其中，并当作周期性系统（PBC）来处理。\n\n具体做法是确保晶胞边长足够大，以避免分子与其镜像相互作用。推荐的晶胞尺寸计算公式为：\nx_len = (max_x - min_x) + max(twobody_cutoff, threebody_cutoff) * 2\ny_len = (max_y - min_y) + max(twobody_cutoff, threebody_cutoff) * 2\nz_len = (max_z - min_z) + max(twobody_cutoff, threebody_cutoff) * 2\n\n其中 max_x\u002Fmin_x 等是分子坐标的最大\u002F最小值，cutoff 是模型的截断半径。这种方法比直接使用 pbc=False 更可靠。","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fissues\u002F82",{"id":130,"question_zh":131,"answer_zh":132,"source_url":133},33422,"导入 MatterSim 时遇到 \"ModuleNotFoundError: No module named 'mattersim.forcefield'; 'mattersim' is not a package\" 错误怎么办？","这通常是因为用户将自己的测试脚本命名为了 `mattersim.py`，导致 Python 导入机制冲突（优先导入了当前目录下的脚本而不是安装的包）。\n\n解决方法：\n1. 检查当前目录下是否有名为 `mattersim.py` 的文件。\n2. 如果有，请将其重命名为其他名称（例如 `test_mattersim.py`）。\n3. 删除同目录下可能生成的 `mattersim.pyc` 文件或 `__pycache__` 文件夹。\n4. 重新运行脚本即可正常导入。","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fissues\u002F104",{"id":135,"question_zh":136,"answer_zh":137,"source_url":138},33423,"运行 MatterSim 最小测试代码时，在 pymatgen 较新版本中报错，提示与 `find_points_in_spheres` 或数据类型有关，如何修复？","该问题是由于 pymatgen 2024.8.9 及更高版本中 `find_points_in_spheres` 函数期望 `pbc` 参数类型为 `np.int64`，而 MatterSim 旧代码传递的是普通 `int` 或 `long`。\n\n修复方法：修改 `src\u002Fmattersim\u002Fdatasets\u002Futils\u002Fconvertor.py` 文件（约第 117 行），将:\n`pbc_ = np.array(structure.pbc, dtype=int)`\n改为:\n`pbc_ = np.array(structure.pbc, dtype=np.int64)`\n\n或者升级 MatterSim 到已包含此修复的最新版本。","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fissues\u002F102",{"id":140,"question_zh":141,"answer_zh":142,"source_url":143},33424,"MatterSim 是否支持 NumPy 2.0 及以上版本？","是的，MatterSim 已经通过 PR #120 修复了对 NumPy >= 2.0 的支持。如果您仍遇到兼容性问题，请确保升级到包含该修复的最新版本。这一更新也解决了 MatterSim 在 conda-forge 上发布 Python 3.13 兼容包的阻塞问题，并使其能在依赖 NumPy 2.0 的平台（如 MLIP Playground）上正常运行。","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fissues\u002F116",{"id":145,"question_zh":146,"answer_zh":147,"source_url":128},33425,"如何处理 MatterSim 在处理非 PBC 结构时晶胞过大导致加载速度慢的问题？","如果自动生成的超胞过大导致数据加载缓慢，可以手动优化晶胞尺寸的计算逻辑。不要简单地乘以一个大倍数（如 10 倍），而是基于原子坐标范围和模型截断半径动态计算。\n\n推荐代码修改如下：\n原代码（可能导致过大）：\nx_len = max((max_x - min_x) * 10, 1000)\n\n修改后（更紧凑且安全）：\nx_len = (max_x - min_x) + max(self.twobody_cutoff, self.threebody_cutoff) * 2\n\n这样既能保证分子与镜像不相互作用，又能显著减小晶胞体积，加快加载速度，且在 ANI1x 数据集上的测试结果显示能量和力的误差几乎没有变化。",{"id":149,"question_zh":150,"answer_zh":151,"source_url":133},33426,"在 WSL2 (Windows Subsystem for Linux) 环境下安装和运行 MatterSim 需要注意什么？","在 WSL2 上运行 MatterSim 本身是支持的，但需注意以下几点以避免常见错误：\n1. 确保没有文件名冲突（如脚本名为 mattersim.py）。\n2. 推荐使用 conda 环境安装，并确保环境激活正确。\n3. 如果从源码安装，请先克隆仓库：`git clone git@github.com:microsoft\u002Fmattersim.git`，然后使用 `conda env create -f environment.yaml` 创建环境。\n4. 若遇到导入错误，优先检查上述文件名冲突问题，这是 WSL2 用户最常遇到的陷阱。",[153,158,163,168,173,178,183,188,193,198,203,208,213,218,223,228,233,238,243,248],{"id":154,"version":155,"summary_zh":156,"released_at":157},255632,"v1.2.2","## 变更内容\n* 由 @danielzuegner 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F143 中修复了应力导入问题\n\n## 新贡献者\n* @danielzuegner 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F143 中完成了首次贡献\n\n**完整变更日志**: https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fcompare\u002Fv1.2.1...v1.2.2","2026-04-05T17:22:45",{"id":159,"version":160,"summary_zh":161,"released_at":162},255633,"v1.2.1","## 变更内容\n* 修复 ASE 3.27 兼容性：由 @yanghan234 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F139 中更新了 Filter 导入和优化器 API。\n* 修复：由 @yanghan234 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F141 中修复了使用 setuptools≥82 时的导入失败问题。\n\n\n**完整变更日志**：https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fcompare\u002Fv1.2.0...v1.2.1","2026-02-25T13:38:13",{"id":164,"version":165,"summary_zh":166,"released_at":167},255634,"v1.2.0","## 🚀 功能与改进\n\n- **BatchRelaxer**: 添加了 `max_n_steps` 参数，以提高灵活性 ([#90](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F90), [@ClaudioZeni](https:\u002F\u002Fgithub.com\u002FClaudioZeni))\n- **基准数据集**: 发布了内部整理的基准数据集 ([#105](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F105), [@yanghan234](https:\u002F\u002Fgithub.com\u002Fyanghan234))\n- **打包**:\n  - 更新了打包流程，使用临时目录 ([#109](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F109), [@bi-ran](https:\u002F\u002Fgithub.com\u002Fbi-ran))\n  - 打包现在在上下文管理器中运行，以提高可靠性 ([#110](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F110), [@bi-ran](https:\u002F\u002Fgithub.com\u002Fbi-ran))\n\n## 🐛 错误修复\n\n- 修复了 `finetune` 中有效路径的处理问题 ([#106](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F106), [@ZeroKnighting](https:\u002F\u002Fgithub.com\u002FZeroKnighting))\n- 修复了 `nsteps` 无法正常工作的问题 ([#101](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F101), [@ClaudioZeni](https:\u002F\u002Fgithub.com\u002FClaudioZeni))\n\n## 🛠️ 维护与 CI\n\n- 升级到 **NumPy 2** ([#120](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F120), [@CompRhys](https:\u002F\u002Fgithub.com\u002FCompRhys))\n- 停止对 **Python 3.9** 的支持 ([#121](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F121), [@yanghan234](https:\u002F\u002Fgithub.com\u002Fyanghan234))\n- CI 改进:\n  - 更新了 `cibuildwheel` 配置 ([#122](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F122), [@yanghan234](https:\u002F\u002Fgithub.com\u002Fyanghan234))\n  - 从 `cibuildwheel` 中移除了 Python 3.9 版本 ([#123](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F123), [@yanghan234](https:\u002F\u002Fgithub.com\u002Fyanghan234))\n\n## 🧹 其他变更\n\n- 修复了拼写错误 ([#114](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F114), [@co63oc](https:\u002F\u002Fgithub.com\u002Fco63oc))\n\n## 🙌 新贡献者\n\n- [@co63oc](https:\u002F\u002Fgithub.com\u002Fco63oc) 完成了首次贡献 ([#114](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F114))\n- [@CompRhys](https:\u002F\u002Fgithub.com\u002FCompRhys) 完成了首次贡献 ([#120](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F120))\n\n**完整变更日志**: [v1.1.2...v1.2.0](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fcompare\u002Fv1.1.2...v1.2.0)\n\n---","2025-07-06T04:39:46",{"id":169,"version":170,"summary_zh":171,"released_at":172},255635,"v1.1.2","## 变更内容\n* 功能：@yanghan234 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F92 中向命令行添加了 `taut` 选项\n* 解除对 `torch\u003C2.5.0` 的版本锁定：@janosh 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F89 中完成\n* CI：@yanghan234 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F93 和 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F94 中更新了 `python-publish.yaml`\n\n**完整变更日志**：https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fcompare\u002Fv1.1.1...v1.1.2\n\n## 变更内容\n* 功能：@yanghan234 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F92 中向命令行添加了 `taut` 选项\n* 解除对 `torch\u003C2.5.0` 的版本锁定：@janosh 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F89 中完成\n* CI：@yanghan234 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F93 和 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F94 中更新了 `python-publish.yaml`\n\n**完整变更日志**：https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fcompare\u002Fv1.1.1...v1.1.2","2025-02-21T05:29:44",{"id":174,"version":175,"summary_zh":176,"released_at":177},255636,"v1.1.2rc3","## 变更内容\n* 功能：由 @yanghan234 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F92 中为命令行添加 `taut` 选项\n* 解除对 `torch\u003C2.5.0` 的版本锁定，由 @janosh 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F89 中完成\n* CI：由 @yanghan234 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F93 中更新 `python-publish.yaml`\n* CI：由 @yanghan234 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F94 中再次更新 `python-publish.yaml`\n\n\n**完整变更日志**：https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fcompare\u002Fv1.1.1...v1.1.2rc3","2025-02-20T15:53:13",{"id":179,"version":180,"summary_zh":181,"released_at":182},255637,"v1.1.2rc2","## 变更内容\n* 功能新增：@yanghan234 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F92 中为命令行添加了 `taut` 选项\n* 依赖更新：@janosh 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F89 中解绑了 `torch\u003C2.5.0` 的版本约束\n* CI 更新：@yanghan234 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F93 中更新了 `python-publish.yaml` 文件\n\n\n**完整变更日志**：https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fcompare\u002Fv1.1.1...v1.1.2rc2","2025-02-20T15:05:12",{"id":184,"version":185,"summary_zh":186,"released_at":187},255638,"v1.1.2rc1","## 变更内容\n* 功能新增：@yanghan234 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F92 中为命令行添加了 `taut` 选项。\n* 依赖更新：@janosh 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F89 中解绑了 `torch\u003C2.5.0` 的版本约束。\n\n\n**完整变更日志**：https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fcompare\u002Fv1.1.1...v1.1.2rc1","2025-02-20T14:34:41",{"id":189,"version":190,"summary_zh":191,"released_at":192},255639,"v1.1.1","## 变更内容\n* mlflow 打包脚本及兼容性改进，由 @bi-ran 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F85 中完成\n* 为 mlflow 锁定 mattersim 依赖版本，由 @bi-ran 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F86 中完成\n* 功能：新增关闭声子谱和声子态密度计算的选项，由 @mbagheri20 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F87 中完成\n* 修复：针对无周期性边界条件的结构，缩小显示框尺寸，由 @anyangml 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F84 中完成\n\n## 新贡献者\n* @bi-ran 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F85 中完成了首次贡献\n* @mbagheri20 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F87 中完成了首次贡献\n* @anyangml 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F84 中完成了首次贡献\n\n**完整变更日志**：https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fcompare\u002Fv1.1.0...v1.1.1","2025-02-05T05:20:40",{"id":194,"version":195,"summary_zh":196,"released_at":197},255640,"v1.1.0","## 变更内容\n\n### 新功能\n- 🔥  **为 MatterSim 添加微调方法**，由 [@brian-xue](https:\u002F\u002Fgithub.com\u002Fbrian-xue) 在 [#68](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F68) 中以及 [@ZeroKnighting](https:\u002F\u002Fgithub.com\u002FZeroKnighting) 在 [#71](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F71) 中实现\n- 🔥 **添加 CLI 脚本，以便在命令行中快速调用计算**，由 [@yanghan234](https:\u002F\u002Fgithub.com\u002Fyanghan234) 在 [#69](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F69) 中实现\n- 🔥  **添加批量弛豫脚本**，由 [@ClaudioZeni](https:\u002F\u002Fgithub.com\u002FClaudioZeni) 在 [#73](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F73) 中实现\n\n### 文档\n- 添加微调相关文档，由 [@ZeroKnighting](https:\u002F\u002Fgithub.com\u002FZeroKnighting) 在 [#78](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F78) 中完成\n- 更新 `finetune.rst` 文件，由 [@yanghan234](https:\u002F\u002Fgithub.com\u002Fyanghan234) 在 [#80](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F80) 中完成\n- 添加批量弛豫相关文档，由 [@ClaudioZeni](https:\u002F\u002Fgithub.com\u002FClaudioZeni) 在 [#75](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F75) 中完成\n\n### 修复\n- 修复 `calculate` 函数中的截止频率自动检测问题，由 [@ZeroKnighting](https:\u002F\u002Fgithub.com\u002FZeroKnighting) 在 [#62](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F62) 中完成\n- 修复使用 CPU 设备时的微调错误，由 [@ZeroKnighting](https:\u002F\u002Fgithub.com\u002FZeroKnighting) 在 [#76](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F76) 中完成\n- 更新 `phonon.py` 文件以保存声子本征矢，由 [@yanghan234](https:\u002F\u002Fgithub.com\u002Fyanghan234) 在 [#77](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F77) 中完成\n\n### 依赖项\n- 更新依赖项，由 [@yanghan234](https:\u002F\u002Fgithub.com\u002Fyanghan234) 在 [#67](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F67) 中完成\n- 移除导致生态系统冲突的 `emmet-core` 降级锁定，由 [@janosh](https:\u002F\u002Fgithub.com\u002Fjanosh) 在 [#61](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F61) 中完成\n\n### 持续集成\n- 更新工作流，由 [@yanghan234](https:\u002F\u002Fgithub.com\u002Fyanghan234) 在 [#72](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F72) 中完成\n\n\n**完整变更日志**: https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fcompare\u002Fv1.0.0...v1.1.0","2025-01-15T05:52:53",{"id":199,"version":200,"summary_zh":201,"released_at":202},255641,"v1.0.1rc1","## 变更内容\n* 修复：在计算中自动检测截断，由 @ZeroKnighting 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F62 中完成\n* 移除导致生态冲突的 emmet-core 降级依赖，由 @janosh 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F61 中完成\n\n## 新贡献者\n* @janosh 在 https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F61 中完成了首次贡献\n\n**完整变更日志**：https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fcompare\u002Fv1.0.0...v1.0.1rc1","2024-12-23T13:13:03",{"id":204,"version":205,"summary_zh":206,"released_at":207},255642,"v1.0.0","## What's Changed\r\n\r\n### Refactors\r\n- Update the method to load checkpoints by @yanghan234 in [#25](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F25)\r\n- Change the location of checkpoints by @yanghan234 in [#54](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F54)\r\n- Update MatterSimCalculator and Potential by @ZeroKnighting in [#37](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F37)\r\n\r\n### Fixes\r\n- Restore DeepCalculator for compatibility by @yanghan234 in [#29](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F29)\r\n- Missing device argument for CPU by @yanghan234 in [#30](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F30)\r\n- Non-PBC handle by @ZeroKnighting in [#50](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F50)\r\n- Cell detection for zero lattice length by @ZeroKnighting in [#53](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F53)\r\n- Potential load method by @ZeroKnighting in [#64](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F64)\r\n\r\n### Documentation\r\n- Update installation instructions by @yanghan234 in [#27](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F27)\r\n- Update documentation by @yanghan234 in [#31](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F31)\r\n- Update README.md multiple times by @yanghan234 in [#32](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F32), [#38](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F38), [#40](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F40), [#42](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F42), [#45](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F57)\r\n- Update docs and fix bugs by @yanghan234 in [#35](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F35)\r\n- Add information on changing checkpoints by @yanghan234 in [#39](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F39)\r\n- Mention AQE in docs by @yanghan234 in [#58](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F58)\r\n- Update picture by @ZeroKnighting in [#34](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F34)\r\n\r\n### Build and CI\r\n- Clean up environment.yaml and reorganize pyproject.toml by @yanghan234 in [#28](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F28)\r\n- Create python-publish.yml by @yanghan234 in [#33](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F33)\r\n- Update python-publish.yaml to build sdist by @yanghan234 in [#43](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F43)\r\n- Update python-publish.yaml and not build sdist by @yanghan234 in [#44](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F44)\r\n- Only deploy pages on push and not pull request by @awvwgk in [#47](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F47)\r\n- Export version information for setuptools_scm in git archives by @awvwgk in [#46](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F46)\r\n- Update build CI to incorporate multiple OS and multiple Python versions by @yanghan234 in [#55](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F55)\r\n- Fix version label for sdist by @yanghan234 in [#56](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F56)\r\n\r\n### Cleanup\r\n- Clean calculator by @ZeroKnighting in [#26](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F26)\r\n\r\n## New Contributors\r\n- @ZeroKnighting made their first contribution in [#26](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F26)\r\n- @awvwgk made their first contribution in [#47](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F47)\r\n\r\n**Full Changelog**: [v1.0.0-rc1...v1.0.0](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fcompare\u002Fv1.0.0-rc1...v1.0.0)","2024-12-18T02:32:38",{"id":209,"version":210,"summary_zh":211,"released_at":212},255643,"v1.0.0-rc10","## What's Changed\r\n* doc: mention AQE in docs by @yanghan234 in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F58\r\n* fix: Potential load method by @ZeroKnighting in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F64\r\n\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fcompare\u002Fv1.0.0-rc9...v1.0.0-rc10","2024-12-16T15:36:26",{"id":214,"version":215,"summary_zh":216,"released_at":217},255644,"v1.0.0-rc9","## What's Changed\r\n* Only deploy pages on push and not pull request by @awvwgk in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F47\r\n* Export version information for setuptools_scm in git archives by @awvwgk in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F46\r\n* fix: non-pbc handle by @ZeroKnighting in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F50\r\n* fix: cell dectection for zero lattice length by @ZeroKnighting in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F53\r\n* refactor: change the location of checkpoints  by @yanghan234 in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F54\r\n* ci: update build ci to incorporate multiple os and multiple python versions by @yanghan234 in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F55\r\n* ci: fix version label for sdist by @yanghan234 in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F56\r\n* doc: update Readme by @yanghan234 in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F57\r\n\r\n## New Contributors\r\n* @awvwgk made their first contribution in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F47\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fcompare\u002Fv1.0.0-rc8...v1.0.0-rc9","2024-12-05T13:46:08",{"id":219,"version":220,"summary_zh":221,"released_at":222},255645,"v1.0.0-rc8","## What's Changed\r\n* doc: Update README.md by @yanghan234 in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F45\r\n\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fcompare\u002Fv1.0.0-rc7...v1.0.0-rc8","2024-12-03T15:14:23",{"id":224,"version":225,"summary_zh":226,"released_at":227},255646,"v1.0.0-rc7","## What's Changed\r\n* ci: Update python-publish.yaml to build sdist by @yanghan234 in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F43\r\n* Update python-publish.yaml and not build sdist by @yanghan234 in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F44\r\n\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fcompare\u002Fv1.0.0-rc6...v1.0.0-rc7","2024-12-03T14:44:02",{"id":229,"version":230,"summary_zh":231,"released_at":232},255647,"v1.0.0-rc6","## What's Changed\r\n* doc: update docs and fix bugs by @yanghan234 in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F35\r\n* refactor: update MatterSimCalculator and Potential by @ZeroKnighting in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F37\r\n* doc: Update README.md by @yanghan234 in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F38\r\n* doc: add information on changing checkpoints by @yanghan234 in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F39\r\n* doc: Update README.md by @yanghan234 in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F40\r\n* doc: update documentations by @yanghan234 in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F41\r\n* doc: Update README.md by @yanghan234 in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F42\r\n\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fcompare\u002Fv1.0.0-rc5...v1.0.0-rc6","2024-12-03T13:50:40",{"id":234,"version":235,"summary_zh":236,"released_at":237},255648,"v1.0.0-rc5","## What's Changed\r\n* doc: update documentation by @yanghan234 in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F31\r\n* doc: Update README.md by @yanghan234 in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F32\r\n* update picture by @ZeroKnighting in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F34\r\n* ci: Create python-publish.yml by @yanghan234 in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F33\r\n\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fcompare\u002Fv1.0.0-rc4...v1.0.0-rc5","2024-12-03T02:34:23",{"id":239,"version":240,"summary_zh":241,"released_at":242},255649,"v1.0.0-rc4","## What's Changed\r\n* fix: missing device argument for cpu by @yanghan234 in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F30\r\n\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fcompare\u002Fv1.0.0-rc3...v1.0.0-rc4","2024-12-02T07:35:04",{"id":244,"version":245,"summary_zh":246,"released_at":247},255650,"v1.0.0-rc3","## What's Changed\r\n* Doc: update installation instruction by @yanghan234 in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F27\r\n* build: clean up environment.yaml and reorganize pyproject.toml by @yanghan234 in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F28\r\n* fix: restore DeepCalculator for compatibility by @yanghan234 in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F29\r\n\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fcompare\u002Fv1.0.0-rc2...v1.0.0-rc3","2024-12-02T04:26:33",{"id":249,"version":250,"summary_zh":251,"released_at":252},255651,"v1.0.0-rc2","## What's Changed\r\n* Refactor: Update the method to load checkpoints by @yanghan234 in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F25\r\n* clean calculator by @ZeroKnighting in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F26\r\n\r\n## New Contributors\r\n* @ZeroKnighting made their first contribution in https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fpull\u002F26\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmattersim\u002Fcompare\u002Fv1.0.0-rc1...v1.0.0-rc2","2024-11-29T04:47:09"]