[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-sammcj--gollama":3,"tool-sammcj--gollama":64},[4,17,27,35,43,56],{"id":5,"name":6,"github_repo":7,"description_zh":8,"stars":9,"difficulty_score":10,"last_commit_at":11,"category_tags":12,"status":16},3808,"stable-diffusion-webui","AUTOMATIC1111\u002Fstable-diffusion-webui","stable-diffusion-webui 是一个基于 Gradio 构建的网页版操作界面，旨在让用户能够轻松地在本地运行和使用强大的 Stable Diffusion 图像生成模型。它解决了原始模型依赖命令行、操作门槛高且功能分散的痛点，将复杂的 AI 绘图流程整合进一个直观易用的图形化平台。\n\n无论是希望快速上手的普通创作者、需要精细控制画面细节的设计师，还是想要深入探索模型潜力的开发者与研究人员，都能从中获益。其核心亮点在于极高的功能丰富度：不仅支持文生图、图生图、局部重绘（Inpainting）和外绘（Outpainting）等基础模式，还独创了注意力机制调整、提示词矩阵、负向提示词以及“高清修复”等高级功能。此外，它内置了 GFPGAN 和 CodeFormer 等人脸修复工具，支持多种神经网络放大算法，并允许用户通过插件系统无限扩展能力。即使是显存有限的设备，stable-diffusion-webui 也提供了相应的优化选项，让高质量的 AI 艺术创作变得触手可及。",162132,3,"2026-04-05T11:01:52",[13,14,15],"开发框架","图像","Agent","ready",{"id":18,"name":19,"github_repo":20,"description_zh":21,"stars":22,"difficulty_score":23,"last_commit_at":24,"category_tags":25,"status":16},1381,"everything-claude-code","affaan-m\u002Feverything-claude-code","everything-claude-code 是一套专为 AI 编程助手（如 Claude Code、Codex、Cursor 等）打造的高性能优化系统。它不仅仅是一组配置文件，而是一个经过长期实战打磨的完整框架，旨在解决 AI 代理在实际开发中面临的效率低下、记忆丢失、安全隐患及缺乏持续学习能力等核心痛点。\n\n通过引入技能模块化、直觉增强、记忆持久化机制以及内置的安全扫描功能，everything-claude-code 能显著提升 AI 在复杂任务中的表现，帮助开发者构建更稳定、更智能的生产级 AI 代理。其独特的“研究优先”开发理念和针对 Token 消耗的优化策略，使得模型响应更快、成本更低，同时有效防御潜在的攻击向量。\n\n这套工具特别适合软件开发者、AI 研究人员以及希望深度定制 AI 工作流的技术团队使用。无论您是在构建大型代码库，还是需要 AI 协助进行安全审计与自动化测试，everything-claude-code 都能提供强大的底层支持。作为一个曾荣获 Anthropic 黑客大奖的开源项目，它融合了多语言支持与丰富的实战钩子（hooks），让 AI 真正成长为懂上",138956,2,"2026-04-05T11:33:21",[13,15,26],"语言模型",{"id":28,"name":29,"github_repo":30,"description_zh":31,"stars":32,"difficulty_score":23,"last_commit_at":33,"category_tags":34,"status":16},2271,"ComfyUI","Comfy-Org\u002FComfyUI","ComfyUI 是一款功能强大且高度模块化的视觉 AI 引擎，专为设计和执行复杂的 Stable Diffusion 图像生成流程而打造。它摒弃了传统的代码编写模式，采用直观的节点式流程图界面，让用户通过连接不同的功能模块即可构建个性化的生成管线。\n\n这一设计巧妙解决了高级 AI 绘图工作流配置复杂、灵活性不足的痛点。用户无需具备编程背景，也能自由组合模型、调整参数并实时预览效果，轻松实现从基础文生图到多步骤高清修复等各类复杂任务。ComfyUI 拥有极佳的兼容性，不仅支持 Windows、macOS 和 Linux 全平台，还广泛适配 NVIDIA、AMD、Intel 及苹果 Silicon 等多种硬件架构，并率先支持 SDXL、Flux、SD3 等前沿模型。\n\n无论是希望深入探索算法潜力的研究人员和开发者，还是追求极致创作自由度的设计师与资深 AI 绘画爱好者，ComfyUI 都能提供强大的支持。其独特的模块化架构允许社区不断扩展新功能，使其成为当前最灵活、生态最丰富的开源扩散模型工具之一，帮助用户将创意高效转化为现实。",107662,"2026-04-03T11:11:01",[13,14,15],{"id":36,"name":37,"github_repo":38,"description_zh":39,"stars":40,"difficulty_score":23,"last_commit_at":41,"category_tags":42,"status":16},3704,"NextChat","ChatGPTNextWeb\u002FNextChat","NextChat 是一款轻量且极速的 AI 助手，旨在为用户提供流畅、跨平台的大模型交互体验。它完美解决了用户在多设备间切换时难以保持对话连续性，以及面对众多 AI 模型不知如何统一管理的痛点。无论是日常办公、学习辅助还是创意激发，NextChat 都能让用户随时随地通过网页、iOS、Android、Windows、MacOS 或 Linux 端无缝接入智能服务。\n\n这款工具非常适合普通用户、学生、职场人士以及需要私有化部署的企业团队使用。对于开发者而言，它也提供了便捷的自托管方案，支持一键部署到 Vercel 或 Zeabur 等平台。\n\nNextChat 的核心亮点在于其广泛的模型兼容性，原生支持 Claude、DeepSeek、GPT-4 及 Gemini Pro 等主流大模型，让用户在一个界面即可自由切换不同 AI 能力。此外，它还率先支持 MCP（Model Context Protocol）协议，增强了上下文处理能力。针对企业用户，NextChat 提供专业版解决方案，具备品牌定制、细粒度权限控制、内部知识库整合及安全审计等功能，满足公司对数据隐私和个性化管理的高标准要求。",87618,"2026-04-05T07:20:52",[13,26],{"id":44,"name":45,"github_repo":46,"description_zh":47,"stars":48,"difficulty_score":23,"last_commit_at":49,"category_tags":50,"status":16},2268,"ML-For-Beginners","microsoft\u002FML-For-Beginners","ML-For-Beginners 是由微软推出的一套系统化机器学习入门课程，旨在帮助零基础用户轻松掌握经典机器学习知识。这套课程将学习路径规划为 12 周，包含 26 节精炼课程和 52 道配套测验，内容涵盖从基础概念到实际应用的完整流程，有效解决了初学者面对庞大知识体系时无从下手、缺乏结构化指导的痛点。\n\n无论是希望转型的开发者、需要补充算法背景的研究人员，还是对人工智能充满好奇的普通爱好者，都能从中受益。课程不仅提供了清晰的理论讲解，还强调动手实践，让用户在循序渐进中建立扎实的技能基础。其独特的亮点在于强大的多语言支持，通过自动化机制提供了包括简体中文在内的 50 多种语言版本，极大地降低了全球不同背景用户的学习门槛。此外，项目采用开源协作模式，社区活跃且内容持续更新，确保学习者能获取前沿且准确的技术资讯。如果你正寻找一条清晰、友好且专业的机器学习入门之路，ML-For-Beginners 将是理想的起点。",84991,"2026-04-05T10:45:23",[14,51,52,53,15,54,26,13,55],"数据工具","视频","插件","其他","音频",{"id":57,"name":58,"github_repo":59,"description_zh":60,"stars":61,"difficulty_score":10,"last_commit_at":62,"category_tags":63,"status":16},3128,"ragflow","infiniflow\u002Fragflow","RAGFlow 是一款领先的开源检索增强生成（RAG）引擎，旨在为大语言模型构建更精准、可靠的上下文层。它巧妙地将前沿的 RAG 技术与智能体（Agent）能力相结合，不仅支持从各类文档中高效提取知识，还能让模型基于这些知识进行逻辑推理和任务执行。\n\n在大模型应用中，幻觉问题和知识滞后是常见痛点。RAGFlow 通过深度解析复杂文档结构（如表格、图表及混合排版），显著提升了信息检索的准确度，从而有效减少模型“胡编乱造”的现象，确保回答既有据可依又具备时效性。其内置的智能体机制更进一步，使系统不仅能回答问题，还能自主规划步骤解决复杂问题。\n\n这款工具特别适合开发者、企业技术团队以及 AI 研究人员使用。无论是希望快速搭建私有知识库问答系统，还是致力于探索大模型在垂直领域落地的创新者，都能从中受益。RAGFlow 提供了可视化的工作流编排界面和灵活的 API 接口，既降低了非算法背景用户的上手门槛，也满足了专业开发者对系统深度定制的需求。作为基于 Apache 2.0 协议开源的项目，它正成为连接通用大模型与行业专有知识之间的重要桥梁。",77062,"2026-04-04T04:44:48",[15,14,13,26,54],{"id":65,"github_repo":66,"name":67,"description_en":68,"description_zh":69,"ai_summary_zh":69,"readme_en":70,"readme_zh":71,"quickstart_zh":72,"use_case_zh":73,"hero_image_url":74,"owner_login":75,"owner_name":76,"owner_avatar_url":77,"owner_bio":78,"owner_company":79,"owner_location":80,"owner_email":81,"owner_twitter":82,"owner_website":83,"owner_url":84,"languages":85,"stars":98,"forks":99,"last_commit_at":100,"license":101,"difficulty_score":23,"env_os":102,"env_gpu":103,"env_ram":104,"env_deps":105,"category_tags":111,"github_topics":112,"view_count":23,"oss_zip_url":81,"oss_zip_packed_at":81,"status":16,"created_at":122,"updated_at":123,"faqs":124,"releases":154},3546,"sammcj\u002Fgollama","gollama","Go manage your Ollama models","gollama 是一款专为 macOS 和 Linux 用户打造的命令行管理工具，旨在帮助用户更高效地打理本地运行的 Ollama 大语言模型。面对日益增多的模型文件，手动通过命令查看、删除或整理往往繁琐且容易出错，gollama 通过提供直观的文字用户界面（TUI），让模型管理变得像操作列表一样简单。\n\n这款工具特别适合经常使用 Ollama 的开发者、AI 研究人员以及技术爱好者。它不仅能列出所有可用模型，还能清晰展示模型大小、量化等级、家族类型及修改日期等关键元数据。用户只需通过键盘快捷键，即可轻松完成模型的排序、筛选、运行、卸载、复制、重命名甚至直接编辑 Modelfile 等操作。此外，gollama 还具备估算显存（vRAM）占用和推送模型到仓库等实用功能，极大地提升了清理旧模型和调试环境的效率。\n\n值得一提的是，尽管项目维护节奏有所调整，但其核心的交互式操作体验依然出色，让复杂的模型运维工作变得更加直观可控。如果你希望在终端中以更优雅的方式掌控本地的 AI 模型资源，gollama 是一个值得尝试的得力助手。","# Gollama\n\n![](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_ca908019fb3c.png)\n\nGollama is a macOS \u002F Linux tool for managing Ollama models.\n\nIt provides a TUI (Text User Interface) for listing, inspecting, deleting, copying, and pushing Ollama models.\n\nThe application allows users to interactively select models, sort, filter, edit, run, unload and perform actions on them using hotkeys.\n\n![](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_ff54a5baf200.jpg)\n\n## Table of Contents\n\n- [Gollama](#gollama)\n  - [Table of Contents](#table-of-contents)\n  - [Features](#features)\n  - [Installation](#installation)\n  - [Usage](#usage)\n  - [Configuration](#configuration)\n  - [Installation and build from source](#installation-and-build-from-source)\n  - [Logging](#logging)\n  - [Contributing](#contributing)\n  - [Acknowledgements](#acknowledgements)\n  - [License](#license)\n\n## Features\n\nGollama is a tool for managing Ollama models with an easy-to-use interface.\n\nIt's in active development, so there are some bugs and missing features, however I'm finding it useful for managing my models every day, especially for cleaning up old models.\n\n- List available models\n- Display metadata such as size, quantisation level, model family, and modified date\n- Edit \u002F update a model's Modelfile\n- Sort models by name, size, modification date, quantisation level, family etc\n- Select and delete models\n- Run and unload models\n- Inspect model for additional details\n- Calculate approximate vRAM usage for a model\n- Copy \u002F rename models\n- Push models to a registry\n- Show running models\n- Has some cool bugs\n\nSee also - [ingest](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fingest) for passing directories\u002Frepos of code to markdown formatted for LLMs.\n\n---\n\n### Update [2025-12-02]: Removal of LM Studio linking & Gollama maintenance slowing\n\nAs of the [v2.0.1 release](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Freleases\u002Ftag\u002Fv2.0.1) of Gollama, LM Studio linking will no longer be available.\n\nLinking from\u002Fto LM Studio became more hassle to maintain than it was worth. Ongoing changes to both upstream applications and trying to cater for each users local configuration meant investing too much of my time for a feature I rarely used.\n\nI'm simply not dog-fooding with Ollama enough. This has meant that development has slowed down as I focus on other projects.\n\nI was an early adopter and contributor to Ollama, but the value I got from Ollama has diminished throughout 2025 to the point where I rarely ever use it. For model serving I have mostly moved to llama.cpp running with [llama-swap](https:\u002F\u002Fgithub.com\u002Fmostlygeek\u002Fllama-swap). Llama.cpp has become far more user friendly over the past year, the project is well maintained, easier to configure, with _many_ more features and _significantly_ better performance. For serving models on my laptop I use [LM Studio](https:\u002F\u002Flmstudio.ai) as it provides both MLX models and the standard llama.cpp runtime for GGUF models.\n\n---\n\n## Installation\n\n### go install (recommended)\n\n```shell\ngo install github.com\u002Fsammcj\u002Fgollama\u002Fv2@latest\n```\n\n### curl\n\nI don't recommend this method as it's not as easy to update, but you can use the following command:\n\n```shell\ncurl -sL https:\u002F\u002Fraw.githubusercontent.com\u002Fsammcj\u002Fgollama\u002Frefs\u002Fheads\u002Fmain\u002Fscripts\u002Finstall.sh | bash\n```\n\n### Manually\n\nDownload the most recent release from the [releases page](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Freleases) and extract the binary to a directory in your PATH.\n\ne.g. `zip -d gollama*.zip -d gollama && mv gollama \u002Fusr\u002Flocal\u002Fbin`\n\n### if \"command not found: gollama\"\n\nIf you see this error, add environment variables to `.zshrc` or `.bashrc`.\n\n```shell\necho 'export PATH=$PATH:$HOME\u002Fgo\u002Fbin' >> ~\u002F.zshrc\nsource ~\u002F.zshrc\n```\n\n## Usage\n\nTo run the `gollama` application, use the following command:\n\n```sh\ngollama\n```\n\n_Tip_: I like to alias gollama to `g` for quick access:\n\n```shell\necho \"alias g=gollama\" >> ~\u002F.zshrc\n```\n\n### Key Bindings\n\n- `Space`: Select\n- `Enter`: Run model (Ollama run)\n- `i`: Inspect model\n- `t`: Top (show running models)\n- `D`: Delete model\n- `e`: Edit model\n- `c`: Copy model\n- `U`: Unload all models\n- `p`: Pull an existing model\n- `ctrl+k`: Pull model & preserve user configuration\n- `ctrl+p`: Pull (get) new model\n- `P`: Push model\n- `n`: Sort by name\n- `s`: Sort by size\n- `m`: Sort by modified\n- `k`: Sort by quantisation\n- `f`: Sort by family\n- `B`: Sort by parameter size\n- `r`: Rename model _**(Work in progress)**_\n- `q`: Quit\n\n#### Top\n\nTop (`t`)\n\n![](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_2c0c797b4c8f.jpg)\n\n#### Inspect\n\nInspect (`i`)\n\n![](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_be0f0a4d8515.png)\n\n#### Command-line Options\n\n**Model Management:**\n- `-l`: List all available Ollama models and exit\n- `-s \u003Csearch term>`: Search for models by name\n  - OR operator (`'term1|term2'`) returns models that match either term\n  - AND operator (`'term1&term2'`) returns models that match both terms\n- `-e \u003Cmodel>`: Edit the Modelfile for a model\n- `-u`: Unload all running models\n- `-v`: Print the version and exit\n\n**Configuration:**\n- `-h`, or `--host`: Specify the host for the Ollama API\n- `-H`: Shortcut for `-h http:\u002F\u002Flocalhost:11434` (connect to local Ollama API)\n- `--ollama-dir`: Custom Ollama models directory\n- `--log` or `--log-level`: Override log level (debug, info, warn, error)\n\n**Cleanup:**\n- `--no-cleanup`: Don't cleanup broken symlinks\n\n**vRAM Analysis:**\n- `--vram`: Estimate vRAM usage for a model. Accepts:\n  - Ollama models (e.g. `llama3.1:8b-instruct-q6_K`, `qwen2:14b-q4_0`)\n  - HuggingFace models (e.g. `NousResearch\u002FHermes-2-Theta-Llama-3-8B`)\n  - `--fits`: Available memory in GB for context calculation (e.g. `6` for 6GB)\n  - `--vram-to-nth` or `--context`: Maximum context length to analyze (e.g. `32k` or `128k`)\n  - `--quant`: Override quantisation level (e.g. `Q4_0`, `Q5_K_M`)\n\n##### Simple model listing\n\nGollama can also be called with `-l` to list models without the TUI.\n\n```shell\ngollama -l\n```\n\nList (`gollama -l`):\n\n![](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_1861d98d0b6b.jpg)\n\n##### Edit\n\nGollama can be called with `-e` to edit the Modelfile for a model.\n\n```shell\ngollama -e my-model\n```\n\n##### Search\n\nGollama can be called with `-s` to search for models by name.\n\n```shell\ngollama -s my-model # returns models that contain 'my-model'\n\ngollama -s 'my-model|my-other-model' # returns models that contain either 'my-model' or 'my-other-model'\n\ngollama -s 'my-model&instruct' # returns models that contain both 'my-model' and 'instruct'\n```\n\n##### vRAM Estimation\n\nGollama includes a comprehensive vRAM estimation feature:\n\n- Calculate vRAM usage for a pulled Ollama model (e.g. `my-model:mytag`), or huggingface model ID (e.g. `author\u002Fname`)\n- Determine maximum context length for a given vRAM constraint\n- Find the best quantisation setting for a given vRAM and context constraint\n- Shows estimates for different k\u002Fv cache quantisation options (fp16, q8_0, q4_0)\n- Automatic detection of available CUDA vRAM (**coming soon!**) or system RAM\n\n![](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_099fd731a560.png)\n\nTo estimate (v)RAM usage:\n\n```shell\ngollama --vram llama3.1:8b-instruct-q6_K\n\n📊 VRAM Estimation for Model: llama3.1:8b-instruct-q6_K\n\n| QUANT   | CTX  | BPW | 2K  | 8K              | 16K             | 32K             | 49K             | 64K |\n| ------- | ---- | --- | --- | --------------- | --------------- | --------------- | --------------- |\n| IQ1_S   | 1.56 | 2.2 | 2.8 | 3.7(3.7,3.7)    | 5.5(5.5,5.5)    | 7.3(7.3,7.3)    | 9.1(9.1,9.1)    |\n| IQ2_XXS | 2.06 | 2.6 | 3.3 | 4.3(4.3,4.3)    | 6.1(6.1,6.1)    | 7.9(7.9,7.9)    | 9.8(9.8,9.8)    |\n| IQ2_XS  | 2.31 | 2.9 | 3.6 | 4.5(4.5,4.5)    | 6.4(6.4,6.4)    | 8.2(8.2,8.2)    | 10.1(10.1,10.1) |\n| IQ2_S   | 2.50 | 3.1 | 3.8 | 4.7(4.7,4.7)    | 6.6(6.6,6.6)    | 8.5(8.5,8.5)    | 10.4(10.4,10.4) |\n| IQ2_M   | 2.70 | 3.2 | 4.0 | 4.9(4.9,4.9)    | 6.8(6.8,6.8)    | 8.7(8.7,8.7)    | 10.6(10.6,10.6) |\n| IQ3_XXS | 3.06 | 3.6 | 4.3 | 5.3(5.3,5.3)    | 7.2(7.2,7.2)    | 9.2(9.2,9.2)    | 11.1(11.1,11.1) |\n| IQ3_XS  | 3.30 | 3.8 | 4.5 | 5.5(5.5,5.5)    | 7.5(7.5,7.5)    | 9.5(9.5,9.5)    | 11.4(11.4,11.4) |\n| Q2_K    | 3.35 | 3.9 | 4.6 | 5.6(5.6,5.6)    | 7.6(7.6,7.6)    | 9.5(9.5,9.5)    | 11.5(11.5,11.5) |\n| Q3_K_S  | 3.50 | 4.0 | 4.8 | 5.7(5.7,5.7)    | 7.7(7.7,7.7)    | 9.7(9.7,9.7)    | 11.7(11.7,11.7) |\n| IQ3_S   | 3.50 | 4.0 | 4.8 | 5.7(5.7,5.7)    | 7.7(7.7,7.7)    | 9.7(9.7,9.7)    | 11.7(11.7,11.7) |\n| IQ3_M   | 3.70 | 4.2 | 5.0 | 6.0(6.0,6.0)    | 8.0(8.0,8.0)    | 9.9(9.9,9.9)    | 12.0(12.0,12.0) |\n| Q3_K_M  | 3.91 | 4.4 | 5.2 | 6.2(6.2,6.2)    | 8.2(8.2,8.2)    | 10.2(10.2,10.2) | 12.2(12.2,12.2) |\n| IQ4_XS  | 4.25 | 4.7 | 5.5 | 6.5(6.5,6.5)    | 8.6(8.6,8.6)    | 10.6(10.6,10.6) | 12.7(12.7,12.7) |\n| Q3_K_L  | 4.27 | 4.7 | 5.5 | 6.5(6.5,6.5)    | 8.6(8.6,8.6)    | 10.7(10.7,10.7) | 12.7(12.7,12.7) |\n| IQ4_NL  | 4.50 | 5.0 | 5.7 | 6.8(6.8,6.8)    | 8.9(8.9,8.9)    | 10.9(10.9,10.9) | 13.0(13.0,13.0) |\n| Q4_0    | 4.55 | 5.0 | 5.8 | 6.8(6.8,6.8)    | 8.9(8.9,8.9)    | 11.0(11.0,11.0) | 13.1(13.1,13.1) |\n| Q4_K_S  | 4.58 | 5.0 | 5.8 | 6.9(6.9,6.9)    | 8.9(8.9,8.9)    | 11.0(11.0,11.0) | 13.1(13.1,13.1) |\n| Q4_K_M  | 4.85 | 5.3 | 6.1 | 7.1(7.1,7.1)    | 9.2(9.2,9.2)    | 11.4(11.4,11.4) | 13.5(13.5,13.5) |\n| Q4_K_L  | 4.90 | 5.3 | 6.1 | 7.2(7.2,7.2)    | 9.3(9.3,9.3)    | 11.4(11.4,11.4) | 13.6(13.6,13.6) |\n| Q5_K_S  | 5.54 | 5.9 | 6.8 | 7.8(7.8,7.8)    | 10.0(10.0,10.0) | 12.2(12.2,12.2) | 14.4(14.4,14.4) |\n| Q5_0    | 5.54 | 5.9 | 6.8 | 7.8(7.8,7.8)    | 10.0(10.0,10.0) | 12.2(12.2,12.2) | 14.4(14.4,14.4) |\n| Q5_K_M  | 5.69 | 6.1 | 6.9 | 8.0(8.0,8.0)    | 10.2(10.2,10.2) | 12.4(12.4,12.4) | 14.6(14.6,14.6) |\n| Q5_K_L  | 5.75 | 6.1 | 7.0 | 8.1(8.1,8.1)    | 10.3(10.3,10.3) | 12.5(12.5,12.5) | 14.7(14.7,14.7) |\n| Q6_K    | 6.59 | 7.0 | 8.0 | 9.4(9.4,9.4)    | 12.2(12.2,12.2) | 15.0(15.0,15.0) | 17.8(17.8,17.8) |\n| Q8_0    | 8.50 | 8.8 | 9.9 | 11.4(11.4,11.4) | 14.4(14.4,14.4) | 17.4(17.4,17.4) | 20.3(20.3,20.3) |\n```\n\nTo find the best quantisation type for a given memory constraint (e.g. 6GB) you can provide `--fits \u003Cnumber of GB>`:\n\n```shell\ngollama --vram NousResearch\u002FHermes-2-Theta-Llama-3-8B --fits 6\n\n📊 VRAM Estimation for Model: NousResearch\u002FHermes-2-Theta-Llama-3-8B\n\n| QUANT\u002FCTX | BPW  | 2K  | 8K  | 16K          | 32K           | 49K            | 64K             |\n| --------- | ---- | --- | --- | ------------ | ------------- | -------------- | --------------- |\n| IQ1_S     | 1.56 | 2.4 | 3.8 | 5.7(4.7,4.2) | 9.5(7.5,6.5)  | 13.3(10.3,8.8) | 17.1(13.1,11.1) |\n| IQ2_XXS   | 2.06 | 2.9 | 4.3 | 6.3(5.3,4.8) | 10.1(8.1,7.1) | 13.9(10.9,9.4) | 17.8(13.8,11.8) |\n...\n```\n\nThis will display a table showing vRAM usage for various quantisation types and context sizes.\n\nThe vRAM estimator works by:\n\n1. Fetching the model configuration from Hugging Face (if not cached locally)\n2. Calculating the memory requirements for model parameters, activations, and KV cache\n3. Adjusting calculations based on the specified quantisation settings\n4. Performing binary and linear searches to optimize for context length or quantisation settings\n\nNote: The estimator will attempt to use CUDA vRAM if available, otherwise it will fall back to system RAM for calculations.\n\n## Configuration\n\nGollama uses a JSON configuration file located at `~\u002F.config\u002Fgollama\u002Fconfig.json`. The configuration file includes options for sorting, columns, API keys, log levels, theme etc...\n\nExample configuration:\n\n```json\n{\n  \"default_sort\": \"modified\",\n  \"columns\": [\n    \"Name\",\n    \"Size\",\n    \"Quant\",\n    \"Family\",\n    \"Modified\",\n    \"ID\"\n  ],\n  \"ollama_api_key\": \"\",\n  \"ollama_api_url\": \"http:\u002F\u002Flocalhost:11434\",\n  \"log_level\": \"info\",\n  \"log_file_path\": \"\u002FUsers\u002Fusername\u002F.config\u002Fgollama\u002Fgollama.log\",\n  \"sort_order\": \"Size\",\n  \"strip_string\": \"my-private-registry.internal\u002F\",\n  \"editor\": \"\u002FApplications\u002FVisual Studio Code.app\u002FContents\u002FResources\u002Fapp\u002Fbin\u002Fcode\",\n  \"docker_container\": \"\"\n}\n```\n\n- `strip_string` can be used to remove a prefix from model names as they are displayed in the TUI. This can be useful if you have a common prefix such as a private registry that you want to remove for display purposes.\n- `editor` specifies which editor to use for editing modelfiles when pressing 'e'. If empty, falls back to the `EDITOR` environment variable, then defaults to `vim`. External editors like VS Code are supported and will show a popup interface.\n- `docker_container` - **experimental** - if set, gollama will attempt to perform any run operations inside the specified container.\n- `theme` - **experimental** The name of the theme to use (without .json extension)\n\n## Installation and build from source\n\n1. Clone the repository:\n\n    ```shell\n    git clone https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama.git\n    cd gollama\n    ```\n\n2. Build:\n\n    ```shell\n    go get\n    make build\n    ```\n\n3. Run:\n\n    ```shell\n    .\u002Fgollama\n    ```\n\n### Themes\n\nGollama has basic customisable theme support, themes are stored as JSON files in `~\u002F.config\u002Fgollama\u002Fthemes\u002F`.\nThe active theme can be set via the `theme` setting in your config file (without the .json extension).\n\nDefault themes will be created if they don't exist:\n\n- `default` - Dark theme with neon accents (default)\n- `light-neon` - Light theme with neon accents, suitable for light terminal backgrounds\n\nTo create a custom theme:\n\n1. Create a new JSON file in the themes directory (e.g. `~\u002F.config\u002Fgollama\u002Fthemes\u002Fmy-theme.json`)\n2. Use the following structure:\n\n```json\n{\n  \"name\": \"my-theme\",\n  \"description\": \"My custom theme\",\n  \"colours\": {\n    \"header_foreground\": \"#AA1493\",\n    \"header_border\": \"#BA1B11\",\n    \"selected\": \"#FFFFFF\",\n    ...\n  },\n  \"family\": {\n    \"llama\": \"#FF1493\",\n    \"alpaca\": \"#FF00FF\",\n    ...\n  }\n}\n```\n\nColours can be specified as ANSI colour codes (e.g. \"241\") or hex values (e.g. \"#FF00FF\"). The `family` section defines colours for different model families in the list view.\n\n_Note: Using the VSCode extension ['Color Highlight'](https:\u002F\u002Fmarketplace.visualstudio.com\u002Fitems?itemName=naumovs.color-highlight) makes it easier to find the hex values for colours._\n\n## Logging\n\nLogs can be found in the `gollama.log` which is stored in `$HOME\u002F.config\u002Fgollama\u002Fgollama.log` by default.\n\nThe log level can be set in the configuration file or overridden via command-line:\n\n```shell\n# Override log level for a single command\ngollama -C --log debug\n\n# Or use the long form\ngollama --create-from-lmstudio --log-level debug\n```\n\nAvailable log levels: `debug`, `info`, `warn`, `error`\n\n## Contributing\n\nContributions are welcome!\nPlease fork the repository and create a pull request with your changes.\n\n\u003C!-- readme: contributors -start -->\n\u003Ctable>\n\t\u003Ctbody>\n\t\t\u003Ctr>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fsammcj\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_b568bd6161ef.png\" width=\"50;\" alt=\"sammcj\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>Sam\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FCamsbury\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_f217828e2a37.png\" width=\"50;\" alt=\"Camsbury\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>Cameron Kingsbury\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FKimCookieYa\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_03e91885acf4.png\" width=\"50;\" alt=\"KimCookieYa\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>KimCookieYa\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fmajiayu000\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_368d4192abab.png\" width=\"50;\" alt=\"majiayu000\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>lif\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FDenisBalan\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_35ebde13410c.png\" width=\"50;\" alt=\"DenisBalan\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>Denis Balan\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Ferg\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_88f1ea1cca18.png\" width=\"50;\" alt=\"erg\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>Doug Coleman\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n\t\t\u003C\u002Ftr>\n\t\t\u003Ctr>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FImpact123\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_8149fd3bc8b6.png\" width=\"50;\" alt=\"Impact123\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>Impact\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fjosekasna\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_4894c6707fbe.png\" width=\"50;\" alt=\"josekasna\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>Jose Almaraz\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fjralmaraz\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_ae27ee0878d3.png\" width=\"50;\" alt=\"jralmaraz\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>Jose Roberto Almaraz\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FBr1ght0ne\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_9ab6c26986bf.png\" width=\"50;\" alt=\"Br1ght0ne\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>Oleksii Filonenko\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fsouthwolf\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_db1b45eaa774.png\" width=\"50;\" alt=\"southwolf\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>SouthWolf\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FVigilans\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_9219d703fd20.png\" width=\"50;\" alt=\"Vigilans\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>Vigilans\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n\t\t\u003C\u002Ftr>\n\t\t\u003Ctr>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fagustif\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_2576a5e81fef.png\" width=\"50;\" alt=\"agustif\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>agustif\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fanrgct\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_89f212c8bd80.png\" width=\"50;\" alt=\"anrgct\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>anrgct\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Ffuho\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_7b748b5481f8.png\" width=\"50;\" alt=\"fuho\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>ondrej\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n\t\t\u003C\u002Ftr>\n\t\u003Ctbody>\n\u003C\u002Ftable>\n\u003C!-- readme: contributors -end -->\n\n## Acknowledgements\n\n- [Ollama](https:\u002F\u002Follama.com\u002F)\n- [Llama.cpp](https:\u002F\u002Fgithub.com\u002Fggerganov\u002Fllama.cpp)\n- [Charmbracelet](https:\u002F\u002Fcharm.sh\u002F)\n\nThank you to folks such as Matt Williams, Fahd Mirza and AI Code King for giving this a shot and providing feedback.\n\n[![AI Code King - Easiest & Interactive way to Manage & Run Ollama Models Locally](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_41f86a1d074f.jpg)](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=T4uiTnacyhI)\n[![Matt Williams - My favourite way to run Ollama: Gollama](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_25151fa977cd.jpg)](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=OCXuYm6LKgE)\n[![Fahd Mirza - Gollama - Manage Ollama Models Locally](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_a773f1254b45.jpg)](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=24yqFrQV-4Q)\n\n## License\n\nCopyright © 2024 Sam McLeod\n\nThis project is licensed under the MIT License. See the [LICENSE](LICENSE) file for details.\n","# Gollama\n\n![](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_ca908019fb3c.png)\n\nGollama 是一款用于管理 Ollama 模型的 macOS \u002F Linux 工具。\n\n它提供了一个文本用户界面（TUI），用于列出、检查、删除、复制和推送 Ollama 模型。\n\n该应用程序允许用户通过热键交互式地选择模型、排序、过滤、编辑、运行、卸载以及对模型执行其他操作。\n\n![](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_ff54a5baf200.jpg)\n\n## 目录\n\n- [Gollama](#gollama)\n  - [目录](#table-of-contents)\n  - [特性](#features)\n  - [安装](#installation)\n  - [使用](#usage)\n  - [配置](#configuration)\n  - [从源码安装与构建](#installation-and-build-from-source)\n  - [日志记录](#logging)\n  - [贡献](#contributing)\n  - [致谢](#acknowledgements)\n  - [许可证](#license)\n\n## 特性\n\nGollama 是一款具有易用界面的 Ollama 模型管理工具。\n\n目前仍在积极开发中，因此可能存在一些 bug 和缺失的功能。不过，我每天都在使用它来管理我的模型，尤其是在清理旧模型时非常有用。\n\n- 列出可用的模型\n- 显示元数据，如大小、量化级别、模型家族和修改日期\n- 编辑\u002F更新模型的 Modelfile\n- 按名称、大小、修改日期、量化级别、家族等对模型进行排序\n- 选择并删除模型\n- 运行和卸载模型\n- 检查模型以获取更多详细信息\n- 计算模型的大致 vRAM 使用量\n- 复制\u002F重命名模型\n- 将模型推送到注册表\n- 显示正在运行的模型\n- 有一些有趣的 bug\n\n另请参阅 - [ingest](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fingest)，用于将代码目录\u002F仓库转换为适合 LLM 的 Markdown 格式。\n\n---\n\n### 更新 [2025-12-02]：移除 LM Studio 链接功能及 Gollama 维护放缓\n\n自 Gollama 的 [v2.0.1 版本](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Freleases\u002Ftag\u002Fv2.0.1) 发布以来，LM Studio 链接功能将不再可用。\n\n与 LM Studio 之间的链接维护起来越来越麻烦，得不偿失。由于上游应用程序不断变化，同时还要适应每个用户的本地配置，这让我投入了过多的时间去维护一个我很少使用的功能。\n\n我自己现在也不再频繁使用 Ollama 了。这也导致开发进度放缓，因为我把精力放在了其他项目上。\n\n我曾是 Ollama 的早期采用者和贡献者，但到了 2025 年，我对 Ollama 的价值已经大不如前，几乎不再使用它。对于模型推理服务，我主要转向了使用 llama.cpp，并结合 [llama-swap](https:\u002F\u002Fgithub.com\u002Fmostlygeek\u002Fllama-swap) 来运行。过去一年里，llama.cpp 变得更加易用，项目维护良好，配置更简单，功能也多了很多，性能更是显著提升。而在笔记本电脑上运行模型时，我则使用 [LM Studio](https:\u002F\u002Flmstudio.ai)，因为它既支持 MLX 模型，也支持标准的 llama.cpp 运行时来处理 GGUF 模型。\n\n---\n\n## 安装\n\n### go install（推荐）\n\n```shell\ngo install github.com\u002Fsammcj\u002Fgollama\u002Fv2@latest\n```\n\n### curl\n\n我不建议使用这种方法，因为它不易于更新，但你也可以使用以下命令：\n\n```shell\ncurl -sL https:\u002F\u002Fraw.githubusercontent.com\u002Fsammcj\u002Fgollama\u002Frefs\u002Fheads\u002Fmain\u002Fscripts\u002Finstall.sh | bash\n```\n\n### 手动安装\n\n从 [发布页面](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Freleases) 下载最新版本，然后将二进制文件解压到你的 PATH 中的一个目录。\n\n例如：`zip -d gollama*.zip -d gollama && mv gollama \u002Fusr\u002Flocal\u002Fbin`\n\n### 如果出现“command not found: gollama”\n\n如果看到此错误，请在 `.zshrc` 或 `.bashrc` 中添加环境变量。\n\n```shell\necho 'export PATH=$PATH:$HOME\u002Fgo\u002Fbin' >> ~\u002F.zshrc\nsource ~\u002F.zshrc\n```\n\n## 使用\n\n要运行 `gollama` 应用程序，可以使用以下命令：\n\n```sh\ngollama\n```\n\n_提示_：我喜欢将 gollama 设置为 `g` 的别名，以便快速访问：\n\n```shell\necho \"alias g=gollama\" >> ~\u002F.zshrc\n```\n\n### 快捷键\n\n- `Space`: 选择\n- `Enter`: 运行模型（Ollama run）\n- `i`: 检查模型\n- `t`: Top（显示正在运行的模型）\n- `D`: 删除模型\n- `e`: 编辑模型\n- `c`: 复制模型\n- `U`: 卸载所有模型\n- `p`: 拉取现有模型\n- `ctrl+k`: 拉取模型并保留用户配置\n- `ctrl+p`: 拉取新模型\n- `P`: 推送模型\n- `n`: 按名称排序\n- `s`: 按大小排序\n- `m`: 按修改日期排序\n- `k`: 按量化级别排序\n- `f`: 按家族排序\n- `B`: 按参数量排序\n- `r`: 重命名模型 _**(正在进行中)**_\n- `q`: 退出\n\n#### Top\n\nTop (`t`)\n\n![](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_2c0c797b4c8f.jpg)\n\n#### 检查\n\nInspect (`i`)\n\n![](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_be0f0a4d8515.png)\n\n#### 命令行选项\n\n**模型管理：**\n- `-l`: 列出所有可用的 Ollama 模型并退出\n- `-s \u003Csearch term>`: 按名称搜索模型\n  - OR 运算符（`'term1|term2'`）返回匹配任一术语的模型\n  - AND 运算符（`'term1&term2'`）返回同时匹配两个术语的模型\n- `-e \u003Cmodel>`: 编辑模型的 Modelfile\n- `-u`: 卸载所有正在运行的模型\n- `-v`: 打印版本号并退出\n\n**配置：**\n- `-h` 或 `--host`: 指定 Ollama API 的主机\n- `-H`: `-h http:\u002F\u002Flocalhost:11434` 的快捷方式（连接到本地 Ollama API）\n- `--ollama-dir`: 自定义 Ollama 模型目录\n- `--log` 或 `--log-level`: 覆盖日志级别（debug、info、warn、error）\n\n**清理：**\n- `--no-cleanup`: 不清理损坏的符号链接\n\n**vRAM 分析：**\n- `--vram`: 估算模型的 vRAM 使用量。接受：\n  - Ollama 模型（如 `llama3.1:8b-instruct-q6_K`、`qwen2:14b-q4_0`）\n  - HuggingFace 模型（如 `NousResearch\u002FHermes-2-Theta-Llama-3-8B`）\n  - `--fits`: 可用内存（GB）用于上下文计算（如 `6` 表示 6GB）\n  - `--vram-to-nth` 或 `--context`: 最大上下文长度（如 `32k` 或 `128k`）\n  - `--quant`: 覆盖量化级别（如 `Q4_0`、 `Q5_K_M`）\n\n##### 简单的模型列表\n\nGollama 也可以通过 `-l` 命令调用，以不使用 TUI 的方式列出模型。\n\n```shell\ngollama -l\n```\n\n列表（`gollama -l`）：\n\n![](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_1861d98d0b6b.jpg)\n\n##### 编辑\n\nGollama 可以通过 `-e` 命令调用，以编辑某个模型的 Modelfile。\n\n```shell\ngollama -e my-model\n```\n\n##### 搜索\n\nGollama 可以通过 `-s` 命令调用，以按名称搜索模型。\n\n```shell\ngollama -s my-model # 返回包含 'my-model' 的模型\n\ngollama -s 'my-model|my-other-model' # 返回包含 'my-model' 或 'my-other-model' 的模型\n\ngollama -s 'my-model&instruct' # 返回同时包含 'my-model' 和 'instruct' 的模型\n```\n\n##### vRAM 估计\n\nGollama 包含全面的 vRAM 估计功能：\n\n- 计算已拉取的 Ollama 模型（如 `my-model:mytag`）或 HuggingFace 模型 ID（如 `author\u002Fname`）的 vRAM 使用量\n- 根据给定的 vRAM 限制确定最大上下文长度\n- 在给定的 vRAM 和上下文限制下找到最佳量化设置\n- 显示不同 k\u002Fv 缓存量化选项的估算值（fp16、q8_0、q4_0）\n- 自动检测可用的 CUDA vRAM（即将推出！）或系统 RAM\n\n![](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_099fd731a560.png)\n\n要估算 vRAM 使用量：\n\n```shell\ngollama --vram llama3.1:8b-instruct-q6_K\n\n📊 VRAM Estimation for Model: llama3.1:8b-instruct-q6_K\n\n| 量化   | 上下文 | BPW | 2K  | 8K              | 16K             | 32K             | 49K             | 64K |\n| ------- | ---- | --- | --- | --------------- | --------------- | --------------- | --------------- |\n| IQ1_S   | 1.56 | 2.2 | 2.8 | 3.7(3.7,3.7)    | 5.5(5.5,5.5)    | 7.3(7.3,7.3)    | 9.1(9.1,9.1)    |\n| IQ2_XXS | 2.06 | 2.6 | 3.3 | 4.3(4.3,4.3)    | 6.1(6.1,6.1)    | 7.9(7.9,7.9)    | 9.8(9.8,9.8)    |\n| IQ2_XS  | 2.31 | 2.9 | 3.6 | 4.5(4.5,4.5)    | 6.4(6.4,6.4)    | 8.2(8.2,8.2)    | 10.1(10.1,10.1) |\n| IQ2_S   | 2.50 | 3.1 | 3.8 | 4.7(4.7,4.7)    | 6.6(6.6,6.6)    | 8.5(8.5,8.5)    | 10.4(10.4,10.4) |\n| IQ2_M   | 2.70 | 3.2 | 4.0 | 4.9(4.9,4.9)    | 6.8(6.8,6.8)    | 8.7(8.7,8.7)    | 10.6(10.6,10.6) |\n| IQ3_XXS | 3.06 | 3.6 | 4.3 | 5.3(5.3,5.3)    | 7.2(7.2,7.2)    | 9.2(9.2,9.2)    | 11.1(11.1,11.1) |\n| IQ3_XS  | 3.30 | 3.8 | 4.5 | 5.5(5.5,5.5)    | 7.5(7.5,7.5)    | 9.5(9.5,9.5)    | 11.4(11.4,11.4) |\n| Q2_K    | 3.35 | 3.9 | 4.6 | 5.6(5.6,5.6)    | 7.6(7.6,7.6)    | 9.5(9.5,9.5)    | 11.5(11.5,11.5) |\n| Q3_K_S  | 3.50 | 4.0 | 4.8 | 5.7(5.7,5.7)    | 7.7(7.7,7.7)    | 9.7(9.7,9.7)    | 11.7(11.7,11.7) |\n| IQ3_S   | 3.50 | 4.0 | 4.8 | 5.7(5.7,5.7)    | 7.7(7.7,7.7)    | 9.7(9.7,9.7)    | 11.7(11.7,11.7) |\n| IQ3_M   | 3.70 | 4.2 | 5.0 | 6.0(6.0,6.0)    | 8.0(8.0,8.0)    | 9.9(9.9,9.9)    | 12.0(12.0,12.0) |\n| Q3_K_M  | 3.91 | 4.4 | 5.2 | 6.2(6.2,6.2)    | 8.2(8.2,8.2)    | 10.2(10.2,10.2) | 12.2(12.2,12.2) |\n| IQ4_XS  | 4.25 | 4.7 | 5.5 | 6.5(6.5,6.5)    | 8.6(8.6,8.6)    | 10.6(10.6,10.6) | 12.7(12.7,12.7) |\n| Q3_K_L  | 4.27 | 4.7 | 5.5 | 6.5(6.5,6.5)    | 8.6(8.6,8.6)    | 10.7(10.7,10.7) | 12.7(12.7,12.7) |\n| IQ4_NL  | 4.50 | 5.0 | 5.7 | 6.8(6.8,6.8)    | 8.9(8.9,8.9)    | 10.9(10.9,10.9) | 13.0(13.0,13.0) |\n| Q4_0    | 4.55 | 5.0 | 5.8 | 6.8(6.8,6.8)    | 8.9(8.9,8.9)    | 11.0(11.0,11.0) | 13.1(13.1,13.1) |\n| Q4_K_S  | 4.58 | 5.0 | 5.8 | 6.9(6.9,6.9)    | 8.9(8.9,8.9)    | 11.0(11.0,11.0) | 13.1(13.1,13.1) |\n| Q4_K_M  | 4.85 | 5.3 | 6.1 | 7.1(7.1,7.1)    | 9.2(9.2,9.2)    | 11.4(11.4,11.4) | 13.5(13.5,13.5) |\n| Q4_K_L  | 4.90 | 5.3 | 6.1 | 7.2(7.2,7.2)    | 9.3(9.3,9.3)    | 11.4(11.4,11.4) | 13.6(13.6,13.6) |\n| Q5_K_S  | 5.54 | 5.9 | 6.8 | 7.8(7.8,7.8)    | 10.0(10.0,10.0) | 12.2(12.2,12.2) | 14.4(14.4,14.4) |\n| Q5_0    | 5.54 | 5.9 | 6.8 | 7.8(7.8,7.8)    | 10.0(10.0,10.0) | 12.2(12.2,12.2) | 14.4(14.4,14.4) |\n| Q5_K_M  | 5.69 | 6.1 | 6.9 | 8.0(8.0,8.0)    | 10.2(10.2,10.2) | 12.4(12.4,12.4) | 14.6(14.6,14.6) |\n| Q5_K_L  | 5.75 | 6.1 | 7.0 | 8.1(8.1,8.1)    | 10.3(10.3,10.3) | 12.5(12.5,12.5) | 14.7(14.7,14.7) |\n| Q6_K    | 6.59 | 7.0 | 8.0 | 9.4(9.4,9.4)    | 12.2(12.2,12.2) | 15.0(15.0,15.0) | 17.8(17.8,17.8) |\n| Q8_0    | 8.50 | 8.8 | 9.9 | 11.4(11.4,11.4) | 14.4(14.4,14.4) | 17.4(17.4,17.4) | 20.3(20.3,20.3) |\n\n要为给定的内存限制（例如 6GB）找到最佳量化类型，可以使用 `--fits \u003CGB 数>` 参数：\n\n```shell\ngollama --vram NousResearch\u002FHermes-2-Theta-Llama-3-8B --fits 6\n\n📊 VRAM 预估：模型 - NousResearch\u002FHermes-2-Theta-Llama-3-8B\n\n| 量化\u002F上下文 | BPW  | 2K  | 8K  | 16K          | 32K           | 49K            | 64K             |\n| --------- | ---- | --- | --- | ------------ | ------------- | -------------- | --------------- |\n| IQ1_S     | 1.56 | 2.4 | 3.8 | 5.7(4.7,4.2) | 9.5(7.5,6.5)  | 13.3(10.3,8.8) | 17.1(13.1,11.1) |\n| IQ2_XXS   | 2.06 | 2.9 | 4.3 | 6.3(5.3,4.8) | 10.1(8.1,7.1) | 13.9(10.9,9.4) | 17.8(13.8,11.8) |\n...\n```\n\n这将显示一张表格，列出不同量化类型和上下文长度下的显存占用情况。\n\n显存估算器的工作原理如下：\n\n1. 从 Hugging Face 获取模型配置（如果本地未缓存）。\n2. 计算模型参数、激活值和 KV 缓存所需的内存。\n3. 根据指定的量化设置调整计算结果。\n4. 使用二分搜索和线性搜索优化上下文长度或量化设置。\n\n注意：估算器会优先使用 CUDA 显存，如果没有可用的 CUDA 显存，则会回退到系统内存进行计算。\n\n\n\n## 配置\n\nGollama 使用位于 `~\u002F.config\u002Fgollama\u002Fconfig.json` 的 JSON 配置文件。该配置文件包含排序选项、列设置、API 密钥、日志级别、主题等配置项。\n\n示例配置：\n\n```json\n{\n  \"default_sort\": \"modified\",\n  \"columns\": [\n    \"Name\",\n    \"Size\",\n    \"Quant\",\n    \"Family\",\n    \"Modified\",\n    \"ID\"\n  ],\n  \"ollama_api_key\": \"\",\n  \"ollama_api_url\": \"http:\u002F\u002Flocalhost:11434\",\n  \"log_level\": \"info\",\n  \"log_file_path\": \"\u002FUsers\u002Fusername\u002F.config\u002Fgollama\u002Fgollama.log\",\n  \"sort_order\": \"Size\",\n  \"strip_string\": \"my-private-registry.internal\u002F\",\n  \"editor\": \"\u002FApplications\u002FVisual Studio Code.app\u002FContents\u002FResources\u002Fapp\u002Fbin\u002Fcode\",\n  \"docker_container\": \"\"\n}\n```\n\n- `strip_string` 可用于移除模型名称前缀，以便在 TUI 中更清晰地显示。如果你的模型名称带有私有仓库等常见前缀，可以通过此设置将其移除。\n- `editor` 指定了按下 'e' 键时使用的编辑器。如果为空，则会回退到 `EDITOR` 环境变量，最后默认使用 `vim`。支持外部编辑器（如 VS Code），并会弹出界面供用户编辑。\n- `docker_container` - **实验性** - 如果设置，Gollama 将尝试在指定的容器内执行所有运行操作。\n- `theme` - **实验性** 主题名称（不带 `.json` 后缀）。\n\n## 安装与源码构建\n\n1. 克隆仓库：\n\n    ```shell\n    git clone https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama.git\n    cd gollama\n    ```\n\n2. 构建：\n\n    ```shell\n    go get\n    make build\n    ```\n\n3. 运行：\n\n    ```shell\n    .\u002Fgollama\n    ```\n\n### 主题\n\nGollama 支持基本的自定义主题功能，主题以 JSON 文件形式存储在 `~\u002F.config\u002Fgollama\u002Fthemes\u002F` 目录中。可通过配置文件中的 `theme` 设置来选择当前主题（不带 `.json` 后缀）。\n\n如果默认主题不存在，系统会自动创建：\n\n- `default` - 带霓虹色点缀的暗色主题（默认）\n- `light-neon` - 带霓虹色点缀的浅色主题，适合浅色终端背景。\n\n创建自定义主题的步骤如下：\n\n1. 在 themes 目录中创建一个新的 JSON 文件（例如 `~\u002F.config\u002Fgollama\u002Fthemes\u002Fmy-theme.json`）。\n2. 使用以下结构：\n\n```json\n{\n  \"name\": \"my-theme\",\n  \"description\": \"我的自定义主题\",\n  \"colours\": {\n    \"header_foreground\": \"#AA1493\",\n    \"header_border\": \"#BA1B11\",\n    \"selected\": \"#FFFFFF\",\n    ...\n  },\n  \"family\": {\n    \"llama\": \"#FF1493\",\n    \"alpaca\": \"#FF00FF\",\n    ...\n  }\n}\n```\n\n颜色可以使用 ANSI 颜色代码（如 `241`）或十六进制值（如 `#FF00FF`）指定。`family` 部分用于定义列表视图中不同模型家族的颜色。\n\n_注意：使用 VSCode 扩展插件 ['Color Highlight'](https:\u002F\u002Fmarketplace.visualstudio.com\u002Fitems?itemName=naumovs.color-highlight) 可以更方便地查找颜色的十六进制值。_\n\n## 日志\n\n日志文件位于 `gollama.log`，默认存储在 `$HOME\u002F.config\u002Fgollama\u002Fgollama.log`。\n\n日志级别可以在配置文件中设置，也可以通过命令行覆盖：\n\n```shell\n# 覆盖单个命令的日志级别\ngollama -C --log debug\n\n# 或者使用长格式\ngollama --create-from-lmstudio --log-level debug\n```\n\n可用的日志级别：`debug`、`info`、`warn`、`error`\n\n## 贡献\n\n欢迎贡献！\n请先 fork 仓库，然后创建包含您更改的 pull request。\n\n\u003C!-- readme: contributors -start -->\n\u003Ctable>\n\t\u003Ctbody>\n\t\t\u003Ctr>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fsammcj\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_b568bd6161ef.png\" width=\"50;\" alt=\"sammcj\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>Sam\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FCamsbury\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_f217828e2a37.png\" width=\"50;\" alt=\"Camsbury\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>Cameron Kingsbury\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FKimCookieYa\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_03e91885acf4.png\" width=\"50;\" alt=\"KimCookieYa\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>KimCookieYa\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fmajiayu000\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_368d4192abab.png\" width=\"50;\" alt=\"majiayu000\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>lif\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FDenisBalan\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_35ebde13410c.png\" width=\"50;\" alt=\"DenisBalan\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>Denis Balan\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Ferg\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_88f1ea1cca18.png\" width=\"50;\" alt=\"erg\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>Doug Coleman\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n\t\t\u003C\u002Ftr>\n\t\t\u003Ctr>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FImpact123\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_8149fd3bc8b6.png\" width=\"50;\" alt=\"Impact123\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>Impact\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fjosekasna\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_4894c6707fbe.png\" width=\"50;\" alt=\"josekasna\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>Jose Almaraz\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fjralmaraz\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_ae27ee0878d3.png\" width=\"50;\" alt=\"jralmaraz\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>Jose Roberto Almaraz\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FBr1ght0ne\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_9ab6c26986bf.png\" width=\"50;\" alt=\"Br1ght0ne\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>Oleksii Filonenko\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fsouthwolf\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_db1b45eaa774.png\" width=\"50;\" alt=\"southwolf\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>SouthWolf\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FVigilans\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_9219d703fd20.png\" width=\"50;\" alt=\"Vigilans\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>Vigilans\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n\t\t\u003C\u002Ftr>\n\t\t\u003Ctr>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fagustif\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_2576a5e81fef.png\" width=\"50;\" alt=\"agustif\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>agustif\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fanrgct\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_89f212c8bd80.png\" width=\"50;\" alt=\"anrgct\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>anrgct\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n            \u003Ctd align=\"center\">\n                \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Ffuho\">\n                    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_7b748b5481f8.png\" width=\"50;\" alt=\"fuho\"\u002F>\n                    \u003Cbr \u002F>\n                    \u003Csub>\u003Cb>ondrej\u003C\u002Fb>\u003C\u002Fsub>\n                \u003C\u002Fa>\n            \u003C\u002Ftd>\n\t\t\u003C\u002Ftr>\n\t\u003Ctbody>\n\u003C\u002Ftable>\n\u003C!-- readme: contributors -end -->\n\n## 致谢\n\n- [Ollama](https:\u002F\u002Follama.com\u002F)\n- [Llama.cpp](https:\u002F\u002Fgithub.com\u002Fggerganov\u002Fllama.cpp)\n- [Charmbracelet](https:\u002F\u002Fcharm.sh\u002F)\n\n感谢 Matt Williams、Fahd Mirza 和 AI Code King 等人试用并提供了反馈。\n\n[![AI Code King - 在本地管理和运行 Ollama 模型的最简单且交互式方式](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_41f86a1d074f.jpg)](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=T4uiTnacyhI)\n[![Matt Williams - 我最喜欢的运行 Ollama 的方式：Gollama](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_25151fa977cd.jpg)](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=OCXuYm6LKgE)\n[![Fahd Mirza - Gollama - 在本地管理 Ollama 模型](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_readme_a773f1254b45.jpg)](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=24yqFrQV-4Q)\n\n## 许可证\n\n版权所有 © 2024 Sam McLeod\n\n本项目采用 MIT 许可证。详情请参阅 [LICENSE](LICENSE) 文件。","# Gollama 快速上手指南\n\nGollama 是一款专为 macOS 和 Linux 设计的终端用户界面（TUI）工具，用于高效管理 Ollama 模型。它支持模型的列表、检查、删除、复制、推送、运行及显存（vRAM）估算等功能。\n\n## 环境准备\n\n在开始之前，请确保满足以下系统要求：\n\n*   **操作系统**：macOS 或 Linux\n*   **前置依赖**：\n    *   已安装并运行 **Ollama** 服务。\n    *   已安装 **Go** 语言环境（推荐版本 1.21+，用于通过 `go install` 安装）。\n*   **网络环境**：确保能够访问 GitHub 及 Go 模块仓库（国内用户若遇网络问题，可配置 `GOPROXY`）。\n\n## 安装步骤\n\n推荐使用 `go install` 方式进行安装，便于后续更新。\n\n### 方法一：使用 Go 安装（推荐）\n\n在终端执行以下命令：\n\n```shell\ngo install github.com\u002Fsammcj\u002Fgollama\u002Fv2@latest\n```\n\n**注意**：如果执行 `gollama` 提示 `command not found`，请将 Go 的二进制目录添加到环境变量中：\n\n```shell\necho 'export PATH=$PATH:$HOME\u002Fgo\u002Fbin' >> ~\u002F.zshrc\nsource ~\u002F.zshrc\n```\n*(注：如果你使用的是 bash，请将 `.zshrc` 替换为 `.bashrc`)*\n\n### 方法二：手动安装\n\n1.  前往 [Releases 页面](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Freleases) 下载最新版本的压缩包。\n2.  解压并将二进制文件移动到系统路径中：\n\n```shell\nunzip gollama*.zip\nmv gollama \u002Fusr\u002Flocal\u002Fbin\n```\n\n### 可选：设置别名\n\n为了方便快速启动，建议设置简短别名（例如 `g`）：\n\n```shell\necho \"alias g=gollama\" >> ~\u002F.zshrc\nsource ~\u002F.zshrc\n```\n\n## 基本使用\n\n### 1. 启动交互式界面\n\n直接在终端输入以下命令即可启动图形化终端界面：\n\n```shell\ngollama\n```\n\n**常用快捷键操作：**\n\n| 按键 | 功能说明 |\n| :--- | :--- |\n| `Space` | 选中\u002F取消选中模型 |\n| `Enter` | 运行选中的模型 (`ollama run`) |\n| `i` | 检查模型详细信息 (Inspect) |\n| `t` | 查看正在运行的模型 (Top) |\n| `D` | 删除选中的模型 |\n| `e` | 编辑模型的 Modelfile |\n| `c` | 复制\u002F重命名模型 |\n| `U` | 卸载所有已加载的模型 |\n| `p` | 拉取现有模型 |\n| `ctrl+p` | 拉取新模型 |\n| `P` | 推送模型到注册表 |\n| `n` \u002F `s` \u002F `m` | 分别按名称、大小、修改时间排序 |\n| `q` | 退出程序 |\n\n### 2. 命令行模式\n\n如果不进入交互界面，也可以使用命令行参数直接操作：\n\n**列出所有模型：**\n```shell\ngollama -l\n```\n\n**搜索模型：**\n```shell\n# 搜索包含 'llama3' 的模型\ngollama -s llama3\n\n# 逻辑或搜索 (包含 'llama3' 或 'instruct')\ngollama -s 'llama3|instruct'\n\n# 逻辑与搜索 (同时包含 'llama3' 和 'instruct')\ngollama -s 'llama3&instruct'\n```\n\n**编辑模型配置：**\n```shell\ngollama -e my-model-name\n```\n\n**卸载所有运行中的模型：**\n```shell\ngollama -u\n```\n\n### 3. 显存 (vRAM) 估算\n\nGollama 提供强大的显存估算功能，可帮助判断模型在特定量化等级和上下文长度下的显存占用。\n\n**估算特定模型的显存占用：**\n```shell\ngollama --vram llama3.1:8b-instruct-q6_K\n```\n\n**根据可用显存查找合适的量化版本：**\n假设你只有 6GB 显存，想找出能运行的最佳量化版本：\n```shell\ngollama --vram NousResearch\u002FHermes-2-Theta-Llama-3-8B --fits 6\n```\n\n该命令会输出一个表格，展示不同量化等级（如 Q4_0, Q5_K_M 等）在不同上下文长度（2K, 8K, 16K...）下的显存需求，帮助你做出最佳选择。","一位本地 AI 开发者在 macOS 上长期运行 Ollama 进行模型实验，随着测试迭代，本地积累了数十个不同版本和量化等级的模型文件。\n\n### 没有 gollama 时\n- **清理困难**：删除旧模型需逐个输入冗长的 `ollama rm` 命令，无法批量操作，清理磁盘空间效率极低。\n- **信息黑盒**：难以直观对比模型的大小、量化等级或修改时间，常因记错名称而误删重要模型或重复下载。\n- **调试繁琐**：查看模型详细元数据或估算显存占用时，需手动查阅文档或编写脚本计算，打断开发心流。\n- **管理混乱**：缺乏排序和过滤功能，在模型列表过长时，寻找特定家族（如 Llama3 或 Mistral）的模型如同大海捞针。\n\n### 使用 gollama 后\n- **一键清理**：通过 TUI 界面用空格键多选过期模型，按 `D` 键即可批量删除，瞬间释放数十 GB 存储空间。\n- **全景可视**：列表直接展示大小、量化级别及修改日期，支持按这些维度排序，模型状态一目了然，杜绝误操作。\n- **即时洞察**：选中模型按 `i` 键即刻 inspect 详细信息，并自动计算近似显存占用，快速判断当前硬件能否加载。\n- **高效流转**：利用快捷键轻松复制、重命名模型或编辑 Modelfile，甚至直接推送至 registry，模型迭代流程丝滑顺畅。\n\ngollama 将原本碎片化、命令行式的模型运维工作，转化为直观的交互式管理体验，极大提升了本地大模型开发的效率与安全性。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fsammcj_gollama_ca908019.png","sammcj","Sam","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002Fsammcj_b568bd61.png","AI Engineer, Platform Engineer, Music Geek || @s_mcleod@aus.social || 📝 Words are my own or somebody else's 🖖","@DigIO","Melbourne, Australia",null,"s_mcleod","https:\u002F\u002Fsmcleod.net","https:\u002F\u002Fgithub.com\u002Fsammcj",[86,90,94],{"name":87,"color":88,"percentage":89},"Go","#00ADD8",97.6,{"name":91,"color":92,"percentage":93},"Makefile","#427819",1.7,{"name":95,"color":96,"percentage":97},"Shell","#89e051",0.7,1734,103,"2026-04-04T15:45:14","MIT","macOS, Linux","非必需。工具本身用于管理模型，不直接运行推理。其显存估算功能可自动检测可用的 CUDA 显存，若无 GPU 则使用系统内存进行计算。","未说明（取决于所管理的 Ollama 模型大小及宿主系统需求）",{"notes":106,"python":107,"dependencies":108},"1. 该工具是使用 Go 语言编写的命令行\u002FTUI 应用，不是 Python 项目，因此无需 Python 环境或 PyTorch 等深度学习库。\n2. 必须预先安装并运行 Ollama 服务（默认端口 11434）才能正常使用模型管理功能。\n3. 可通过 'go install' 直接安装，或下载预编译二进制文件。\n4. 配置文件位于 ~\u002F.config\u002Fgollama\u002Fconfig.json。","不需要 (基于 Go 语言开发)",[109,110],"Go (编译环境\u002F运行时)","Ollama (后端服务)",[15,14,13,26],[113,114,115,116,117,118,119,120,121],"ai","gguf","llm","ollama","tui","ggml","linux","macos","models","2026-03-27T02:49:30.150509","2026-04-06T07:14:49.879818",[125,130,134,139,144,149],{"id":126,"question_zh":127,"answer_zh":128,"source_url":129},16246,"为什么在 Mac 上使用 `go install` 安装后提示 \"command not found\"？","这通常不是软件本身的 Bug，而是用户环境或 OS 配置问题。请检查您的 `$PATH` 环境变量是否包含了 Go 的二进制文件目录（通常是 `$HOME\u002Fgo\u002Fbin`）。如果不确定，可以尝试重启终端或手动将该路径添加到 shell 配置文件（如 `.zshrc` 或 `.bash_profile`）中。维护者表示无法复现此问题，建议确认系统配置是否正确。","https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F132",{"id":131,"question_zh":132,"answer_zh":133,"source_url":129},16247,"运行 gollama 时遇到 \"Error fetching models: dial tcp 0.0.0.0:80: connection refused\" 错误怎么办？","这是因为环境变量 `OLLAMA_HOST` 被错误地设置为了 `0.0.0.0`，导致程序尝试连接错误的端口。解决方法是将其显式设置为正确的地址。请在终端执行以下命令：\n`export OLLAMA_HOST=localhost:11434`\n设置完成后再次运行 `.\u002Fgollama` 即可正常工作。",{"id":135,"question_zh":136,"answer_zh":137,"source_url":138},16248,"在 macOS 上使用 `-L` 链接模型时，为什么提示默认路径错误或找不到 LM Studio 模型目录？","macOS 上 LM Studio 的默认模型存储路径已更新为 `~\u002F.lmstudio\u002Fmodels`，而旧版本或默认配置可能仍在查找 `~\u002F.cache\u002Flm-studio\u002Fmodels`。如果您将模型存放在自定义路径或非默认路径，需要使用 `-lm-dir` 参数指定正确的目录。例如：\n`gollama -L -lm-dir \u002FUsers\u002Fyour_name\u002F.lmstudio\u002Fmodels`\n请确保指定的路径确实存在且包含模型文件。","https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F151",{"id":140,"question_zh":141,"answer_zh":142,"source_url":143},16249,"Gollama 在浅色背景（Light Background）的终端中显示效果不佳（文字透明或看不清）如何解决？","这是一个已知的外观问题，维护者已在最新版本中合并了主题（Themes）功能来解决此问题。您可以尝试更新到最新版本以启用新的配色方案。如果您擅长配色调整，也可以贡献自己的浅色主题配置文件给项目，以帮助进一步优化显示效果。","https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F167",{"id":145,"question_zh":146,"answer_zh":147,"source_url":148},16250,"使用 VRAM 估算功能时出现 \"401 Unauthorized\" 错误是什么原因？","该错误通常发生在需要访问 Hugging Face API 获取模型信息但未提供认证令牌时。解决方法是设置 `HUGGINGFACE_TOKEN` 环境变量。请在终端执行：\n`export HUGGINGFACE_TOKEN=your_token_here`\n将 `your_token_here` 替换为您实际的 Hugging Face Token。设置后重新运行命令即可。维护者已在最新版本中改进了对此类错误的处理提示。","https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F87",{"id":150,"question_zh":151,"answer_zh":152,"source_url":153},16251,"Gollama 的 VRAM 估算输出中的三个数值分别代表什么含义？","虽然具体数值的详细定义在讨论中未完全展开，但 VRAM 估算功能旨在计算模型在不同上下文长度（Context Length）下的显存需求。估算逻辑通常包含三部分：模型权重本身的大小（Model Size）、KV Cache 占用的显存（取决于上下文长度和量化位数）以及激活内存（Activation Memory）。用户可以通过编写脚本调用 `gollama -l` 解析输出，结合参数量、量化位数和上下文长度来自动化评估模型是否能放入显存。","https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F117",[155,160,165,170,175,180,185,190,195,200,205,210,215,220,225,230,235,240,245,250],{"id":156,"version":157,"summary_zh":158,"released_at":159},98537,"v2.0.4","### [2.0.4](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fvv2.0.3...v2.0.4) (2025-12-30)\n\n**完整更新日志**: https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fv2.0.3...v2.0.4","2025-12-30T22:31:36",{"id":161,"version":162,"summary_zh":163,"released_at":164},98538,"v2.0.3","### [2.0.3](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fvv2.0.2...v2.0.3)（2025-12-30）\n\n\n### Bug 修复\n\n* 内部包版本，杂项：升级依赖项（[#232](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F232)）（[46255ef](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcommit\u002F46255ef7aec6fb5d8c3711ec3102ecd960bca47e)）\n\n## 变更内容\n* 修复：内部包版本，杂项：升级依赖项，由 @sammcj 在 https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F232 中完成\n* 文档（贡献者）：更新贡献者 README 中的 GitHub Actions 流程，由 @github-actions[bot] 在 https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F229 中完成\n* 文档（贡献者）：更新贡献者 README 中的 GitHub Actions 流程，由 @github-actions[bot] 在 https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F231 中完成\n* 文档（贡献者）：更新贡献者 README 中的 GitHub Actions 流程，由 @github-actions[bot] 在 https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F233 中完成\n\n\n**完整变更日志**：https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fv2.0.2...v2.0.3","2025-12-30T22:16:29",{"id":166,"version":167,"summary_zh":168,"released_at":169},98539,"v2.0.2","### [2.0.2](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fvv2.0.1...v2.0.2) (2025-12-30)\n\n\n### Bug 修复\n\n* 🐞 问题：重命名模型后按下 Ctrl-C 会导致模型被删除（[#228](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F228)）（[597bbf7](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcommit\u002F597bbf7e9a23efcbbeea87908b85cbd9442a665e)），关闭了 [#227](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F227)\n* 功能：使用过滤器时的排序（[#230](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F230)）（[2f17393](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcommit\u002F2f1739340183fa6cfdfe305ff3d7729d157e8ede））\n\n\n### 文档更新\n\n* 更新自述文件 [skip-ci]（[cb1f2d8](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcommit\u002Fcb1f2d87f8b463d6b789d8a13ad7137a83534000)）\n\n## 变更内容\n* 修复：🐞 问题：重命名模型后按下 Ctrl-C 会导致模型被删除，由 @majiayu000 在 https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F228 中修复。\n* 修复：功能：使用过滤器时的排序，由 @majiayu000 在 https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F230 中实现。\n\n## 新贡献者\n* @majiayu000 在 https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F228 中做出了首次贡献。\n\n**完整变更日志**：https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fv2.0.1...v2.0.2","2025-12-30T22:00:59",{"id":171,"version":172,"summary_zh":173,"released_at":174},98540,"v2.0.1","## [2.0.1](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fvv1.37.5...v2.0.1) (2025-12-02)\n\n\n### ⚠ 重大变更\n\n**自本版本 Gollama 起，LM Studio 链接和模型拆分功能将不再可用。**\n\n- 移除所有 LM Studio 链接功能\n- 移除所有模型拆分功能\n\n与 LM Studio 之间的链接维护起来越来越麻烦，得不偿失。由于上游应用不断变化，同时还要适配每位用户的本地配置，这使得该功能占用了我过多的时间，而我自己却很少使用它。\n\n我目前对 Ollama 的实际使用频率已经很低，无法真正做到“自用驱动开发”。这也导致我的开发进度放缓，不得不将更多精力投入到其他项目中。\n\n我曾是 Ollama 的早期用户和贡献者，但随着 2025 年的推移，我对 Ollama 的依赖和价值感逐渐降低，以至于现在几乎不再使用它。对于模型推理服务，我已主要转向使用 llama.cpp，并配合 [llama-swap](https:\u002F\u002Fgithub.com\u002Fmostlygeek\u002Fllama-swap) 运行。过去一年里，llama.cpp 的易用性大幅提升，项目维护良好、配置更简单，功能也更加丰富，性能更是显著提高。而在笔记本上部署模型时，我则选择使用 [LM Studio](https:\u002F\u002Flmstudio.ai)，因为它既支持 MLX 模型，也兼容标准的 llama.cpp 运行时来加载 GGUF 格式的模型。\n\n---\n\n### 重大变更\n\n* 移除所有 LM Studio 链接和模型拆分功能（[#225](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F225)）（[a9224f8](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcommit\u002Fa9224f83b948df940f97218607523ca75e43d38d)），关闭了 [#221](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F221)、[#224](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F224) 和 [#191](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F191)\n\n## 变更内容\n* 重大变更：移除所有 LM Studio 链接和模型拆分功能，由 @sammcj 在 https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F225 中完成\n\n\n**完整变更日志**：https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fv1.37.5...v2.0.1","2025-12-02T09:57:13",{"id":176,"version":177,"summary_zh":178,"released_at":179},98541,"v2.0.0","## [2.0.0](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fvv1.37.5...v2.0.0) (2025-12-02)\n\n\n### ⚠ 重大变更\n\n**自 Gollama 下一个版本起，LM Studio 链接和模型拆分功能将不再可用。**\n\n- 移除所有 LM Studio 链接功能\n- 移除所有模型拆分功能\n\n我决定移除这两个功能，因为维护它们的成本远高于其带来的价值。\n\n### 重大变更\n\n* 移除所有 LM Studio 链接和模型拆分功能（[#225](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F225)）（[a9224f8](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcommit\u002Fa9224f83b948df940f97218607523ca75e43d38d)），关闭了 [#221](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F221)、[#224](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F224) 和 [#191](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F191)\n\n## 变更内容\n* 重大变更：移除所有 LM Studio 链接和模型拆分功能，由 @sammcj 在 https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F225 中完成\n\n\n**完整更新日志**：https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fv1.37.5...v2.0.0","2025-12-02T09:54:48",{"id":181,"version":182,"summary_zh":183,"released_at":184},98542,"v1.37.5","### [1.37.5](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fvv1.37.4...v1.37.5)（2025-11-13）\n\n## 变更内容\n* 添加 Claude Code GitHub 工作流 [skip-ci]，由 @sammcj 在 https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F223 中完成\n\n\n**完整变更日志**：https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fv1.37.4...v1.37.5","2025-11-13T05:39:24",{"id":186,"version":187,"summary_zh":188,"released_at":189},98543,"v1.37.4","### [1.37.4](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fvv1.37.3...v1.37.4) (2025-11-11)\n\n\n### 错误修复\n\n* LMStudio 到 Ollama 模型的创建 ([#222](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F222)) ([c429cc3](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcommit\u002Fc429cc3ae5a82f0d7c8ab04531d66d774043ba60))\n\n## 变更内容\n* 修复：由 @sammcj 在 https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F222 中实现的 LMStudio 到 Ollama 模型的创建问题\n\n\n**完整变更日志**: https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fv1.37.3...v1.37.4","2025-11-11T21:59:35",{"id":191,"version":192,"summary_zh":193,"released_at":194},98544,"v1.37.3","### [1.37.3](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fvv1.37.2...v1.37.3)（2025-09-29）\n\n\n### 错误修复\n\n* **清理：** 在清理时移除空目录（[#218](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F218)）（[1358225](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcommit\u002F1358225cd992713d72bce9237c4d4b30096b3ac3)）\n\n## 变更内容\n* 修复（清理）：由 @sammcj 在 https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F218 中实现，在清理时移除空目录\n\n\n**完整变更日志**：https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fv1.37.2...v1.37.3","2025-09-29T03:25:50",{"id":196,"version":197,"summary_zh":198,"released_at":199},98545,"v1.37.2","### [1.37.2](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fvv1.37.1...v1.37.2)（2025-09-16）\n\n\n### 错误修复\n\n* 修复编辑名称中包含空格的模型的问题 ([ce494ae](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcommit\u002Fce494ae55811f3915811981c9d7b3d89d9aa8703))\n\n**完整更新日志**: https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fv1.37.1...v1.37.2","2025-09-16T04:25:30",{"id":201,"version":202,"summary_zh":203,"released_at":204},98546,"v1.37.1","### [1.37.1](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fvv1.37.0...v1.37.1)（2025-08-25）\n\n\n### Bug 修复\n\n* 改进符号链接清理（[#213](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F213)）（[ee02a25](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcommit\u002Fee02a2542e71033fd80b789c68ace94c50e37531)）\n\n## 变更内容\n* 修复：由 @sammcj 在 https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F213 中改进的符号链接清理\n\n\n**完整变更日志**：https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fv1.37.0...v1.37.1","2025-08-25T00:18:40",{"id":206,"version":207,"summary_zh":208,"released_at":209},98547,"v1.37.0","## [1.37.0](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fvv1.36.1...v1.37.0) (2025-08-24)\n\n\n### Features\n\n* samll go binaries ([25026e5](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcommit\u002F25026e50d66d5c4c92cdc93c096d29fd35660dc9))\n\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fv1.36.1...v1.37.0","2025-08-24T10:25:25",{"id":211,"version":212,"summary_zh":213,"released_at":214},98548,"v1.36.1","### [1.36.1](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fvv1.36.0...v1.36.1) (2025-08-22)\n\n\n### Bug Fixes\n\n* improve symlink cleanup ([#211](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F211)) ([07baa59](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcommit\u002F07baa59e30cd34aa5d28c10c89fbd3c77db7657f))\n\n## What's Changed\n* docs(contributor): contributors readme action update by @github-actions[bot] in https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F210\n* fix: improve symlink cleanup by @sammcj in https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F211\n\n\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fv1.36.0...v1.36.1","2025-08-22T03:04:49",{"id":216,"version":217,"summary_zh":218,"released_at":219},98549,"v1.36.0","## [1.36.0](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fvv1.35.3...v1.36.0) (2025-08-22)\n\n\n### Features\n\n* **linking:** use canonical model name when linking to lm-studio ([#209](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F209)) ([f28b89b](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcommit\u002Ff28b89b7966a53fa7496e9a09696587e325f11d4))\n\n## What's Changed\n* feat(linking): use canonical model name when linking to lm-studio by @Vigilans in https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F209\n\n## New Contributors\n* @Vigilans made their first contribution in https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F209\n\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fv1.35.3...v1.36.0","2025-08-22T01:52:35",{"id":221,"version":222,"summary_zh":223,"released_at":224},98550,"v1.35.3","### [1.35.3](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fvv1.35.2...v1.35.3) (2025-08-10)\n\n\n### Bug Fixes\n\n* **linking:** preserve date\u002Ftime modified when linking to lm-studio ([#206](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F206)) ([d4df10d](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcommit\u002Fd4df10dc7a30ed05cbf92486d164acc1099769a0))\n\n## What's Changed\n* fix(linking): preserve date\u002Ftime modified when linking to lm-studio by @sammcj in https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F206\n\n\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fv1.35.2...v1.35.3","2025-08-10T05:51:25",{"id":226,"version":227,"summary_zh":228,"released_at":229},98551,"v1.35.2","### [1.35.2](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fvv1.35.1...v1.35.2) (2025-08-10)\n\n## What's Changed\n* chore(packages): bump deps by @sammcj in https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F205\n\n\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fv1.35.1...v1.35.2","2025-08-10T05:40:18",{"id":231,"version":232,"summary_zh":233,"released_at":234},98552,"v1.35.1","### [1.35.1](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fvv1.35.0...v1.35.1) (2025-07-28)\n\n## What's Changed\n* chore(packages): bump deps by @sammcj in https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F200\n\n\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fv1.35.0...v1.35.1","2025-07-28T01:01:11",{"id":236,"version":237,"summary_zh":238,"released_at":239},98553,"v1.35.0","## [1.35.0](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fvv1.34.1...v1.35.0) (2025-07-28)\n\n\n### Features\n\n* support configurable text editor for editing modelfiles ([#199](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F199)) ([b4b7755](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcommit\u002Fb4b77555771f5c9401b780519987a309f88f2cd9))\n\n## What's Changed\n* feat: support configurable text editor for editing modelfiles by @sammcj in https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F199\n\n\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fv1.34.1...v1.35.0","2025-07-28T00:29:43",{"id":241,"version":242,"summary_zh":243,"released_at":244},98554,"v1.34.1","### [1.34.1](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fvv1.34.0...v1.34.1) (2025-07-13)\n\n\n### Features\n\n* Link Ollama models from LM Studio models.  ([#198](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F198)) ([8111eaa](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcommit\u002F8111eaaa3b5036ca91cfb972fab0e665f0cac047))\n\n## What's Changed\n* Feat: Link Ollama models from LM Studio models.  by @sammcj in https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F198\n\n\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fv1.34.0...v1.34.1","2025-07-13T08:57:21",{"id":246,"version":247,"summary_zh":248,"released_at":249},98555,"v1.34.0","## [1.34.0](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fvv1.33.2...v1.34.0) (2025-06-12)\n\n\n### Features\n\n* use Ollamas new capabilities API for model info ([#195](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F195)) ([b31ab4b](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcommit\u002Fb31ab4bfd0f19483ae1d3627e8d2653bac87eab2))\n\n## What's Changed\n* docs(contributor): contributors readme action update by @github-actions in https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F190\n* feat: use Ollamas new capabilities API for model info by @sammcj in https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F195\n\n\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fv1.33.2...v1.34.0","2025-06-12T03:41:53",{"id":251,"version":252,"summary_zh":253,"released_at":254},98556,"v1.33.2","### [1.33.2](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fvv1.33.1...v1.33.2) (2025-05-08)\n\n\n### Bug Fixes\n\n* **lmstudio:** include symlinked .gguf and .bin files in ScanModels ([#189](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fissues\u002F189)) ([06ed41a](https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcommit\u002F06ed41ad67b11cb81b48d50c48e4e7ccb1810583))\n\n## What's Changed\n* fix(lmstudio): include symlinked model files in ScanModels & document bidirectional sync by @agustif in https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F189\n\n## New Contributors\n* @agustif made their first contribution in https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fpull\u002F189\n\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fsammcj\u002Fgollama\u002Fcompare\u002Fv1.33.1...v1.33.2","2025-05-08T04:57:49"]