Tangle – 可视化机器学习流水线编辑器
Tangle – Visual ML Pipeline Editor

原始链接: https://tangleml.com/

Tangle 是一个由 Shopify 开发的开源、与平台无关的实验平台,旨在普及机器学习流水线开发。它通过功能强大的拖拽式可视化编辑器,让团队能够构建复杂的机器学习工作流,无需繁琐的配置。 主要功能包括: * **协作设计:** 团队可以使用可重用模块库或自定义组件来克隆、共享和编辑流水线。 * **灵活性:** 作为一种平台无关工具,Tangle 允许用户通过将代码封装在容器中来集成任何语言或框架。 * **高效性:** 基于内容的先进缓存机制可存储中间结果,显著缩短实验时间并降低计算成本。 无论是从零开始构建还是使用示例流水线,Tangle 都为现代机器学习团队提供了一个易于访问的协作环境。您可以在 GitHub 上探索该项目、为其开发做出贡献或查看文档。

Hacker News 最新 | 往日 | 评论 | 提问 | 展示 | 招聘 | 提交 登录 Tangle – 可视化机器学习流水线编辑器 (tangleml.com) 18 分,作者:duck,2 小时前 | 隐藏 | 往日 | 收藏 | 2 条评论 0x457 5 分钟前 | 下一条 [–] 很棒,我也在开发类似的东西。但在编辑器中构建了几个工作流后,我意识到它们基本上都是一样的:获取已标注的数据集,训练以生成模型,然后在其他数据集上使用该模型。 回复 dleeftink 2 小时前 | 上一条 [–] 如果你对 KNIME[0] 感到厌倦了,这看起来是个不错的替代品。[0]: https://www.knime.com/why-visual-workflows 回复 社区准则 | 常见问题 | 列表 | API | 安全 | 法律 | 申请 YC | 联系 搜索:
相关文章

原文

Build ML and data pipelines collaboratively

Tangle is an open source, platform-agnostic experimentation platform with a powerful drag-and-drop editor – no setup required.

Shopify LogoPowered by Shopify

Developed by Shopify ML and software engineers to democratize machine learning pipeline development.

Open SourceOpen Source

Access publicly available code and contribute to the future of visual ML pipeline development.

Powerful Features for ML Teams

Drag & Drop Pipeline Editing

Create complex ML workflows with a powerful visual editor. Connect components, configure arguments, and build pipelines– with or without writing code.

Drag and drop interface showing pipeline components

Build Pipelines Together

Clone, edit and run any pipeline as you see fit. Choose from reusable modules in the component library or add your own.

A list of pipelines runs from many different users

Any Language, Any Framework

Tangle is platform agnostic. Wrap your existing code in a container using the in-app editor.

In-app editor wizard

Advanced Execution Caching

Cache intermediate results to accelerate experimentation. Content-based caching saves significant time and compute costs.

Pipeline execution page

Ready to Build Your First Pipeline?

Start creating machine learning workflows in minutes. Try our example pipelines or dive straight into the documentation.

Get in Touch

Have questions about Tangle? Our team is here to help you get started with visual ML pipeline development.

Open Source

Tangle is built on open-source foundations. Contribute, report issues, or explore the codebase on GitHub.

联系我们 contact @ memedata.com