Skip to content

Commit

Permalink
🚀feat: update what-is-swanlab
Browse files Browse the repository at this point in the history
  • Loading branch information
@Zeyi-Lin
Zeyi-Lin committed Jan 8, 2025
1 parent d202a4e commit 24f032b
Show file tree
Hide file tree
Showing 4 changed files with 33 additions and 30 deletions.
Binary file added assets/integrations.png
View file
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file modified assets/product-swanlab-1.png
View file
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
21 changes: 7 additions & 14 deletions en/guide_cloud/general/what-is-swanlab.md
View file
Original file line number Diff line number Diff line change
@@ -1,16 +1,11 @@
# Welcome to SwanLab

[Official Website](https://dev101.swanlab.cn) · [Github](https://github.com/swanhubx/swanlab) · [Contribution Guide
](https://github.com/SwanHubX/SwanLab/blob/main/CONTRIBUTING.md) · [SwanHub](https://swanhub.co) · [VS Code Extension](https://marketplace.visualstudio.com/items?itemName=SwanLab.swanlab&ssr=false#overview)
[SwanLab.cn](https://swanlab.cn) · [Github](https://github.com/swanhubx/swanlab) · [VSCode Extension](https://marketplace.visualstudio.com/items?itemName=SwanLab.swanlab&ssr=false#overview) · [Quick Start](/en/guide_cloud/general/quick-start.md)

<!-- ![](/assets/swanlab-show.png) -->

![alt text](/assets/product-swanlab-1.png)

::: warning 👋 We've just launched our VS Code extension, try it out!
Search "SwanLab" in the VS Code Extension Marketplace to install.
:::

SwanLab is an open-source, lightweight AI experiment tracking tool that provides a platform for tracking, comparing, and collaborating on experiments, designed to accelerate the R&D efficiency of AI R&D teams by 100 times.

It provides a user-friendly API and a beautiful interface, combined with hyperparameter tracking, metric logging, online collaboration, experiment link sharing, real-time message notification and other functions, allowing you to quickly track ML experiments, visualize the process, and share with peers.
Expand All @@ -37,8 +32,6 @@ We hope that the above information and this guide can help you understand this p

## What can SwanLab do?

![alt text](/assets/how-to-do-swanlab.png)

**1. 📊Experiment metrics and hyperparameter tracking**: Minimal code embedded in your machine learning pipeline to track and record key training metrics
- Free hyperparameter and experiment configuration recording
- **Supported metadata types**: scalar metrics, images, audio, text, etc.
Expand All @@ -54,11 +47,11 @@ We hope that the above information and this guide can help you understand this p

## Where to start

- [Quick Start](/zh/guide_cloud/general/quick-start.md): SwanLab introductory tutorial, master experiment tracking in five minutes!
- [API documentation](/zh/api/api-index.md): Complete API documentation
- [Online support](/zh/guide_cloud/community/online-support.md): Join the community, provide feedback and contact us
- [Self-hosting](/zh/guide_cloud/self_host/offline-board.md): Self-hosting (offline version) usage tutorial
- [Cases](/zh/examples/mnist.md): View cases of SwanLab with various deep learning tasks
- [Quick Start](/en/guide_cloud/general/quick-start.md): SwanLab introductory tutorial, master experiment tracking in five minutes!
- [API documentation](/en/api/api-index.md): Complete API documentation
- [Online support](/en/guide_cloud/community/online-support.md): Join the community, provide feedback and contact us
- [Self-hosting](/en/guide_cloud/self_host/offline-board.md): Self-hosting (offline version) usage tutorial
- [Cases](/en/examples/mnist.md): View cases of SwanLab with various deep learning tasks

## Comparison with familiar products

Expand Down Expand Up @@ -89,7 +82,7 @@ We hope that the above information and this guide can help you understand this p
- **Email Support**: Feedback regarding issues with SwanLab
- Product: <[email protected]>, <[email protected]>(Product Manager Email)

- **WeChat Group and Lark Group**: See [Online Support](/zh/guide_cloud/community/online-support.md)
- **WeChat Group and Lark Group**: See [Online Support](/en/guide_cloud/community/online-support.md)

- **WeChat Official Account**:

Expand Down
42 changes: 26 additions & 16 deletions zh/guide_cloud/general/what-is-swanlab.md
View file
Original file line number Diff line number Diff line change
@@ -1,21 +1,27 @@
# 欢迎使用SwanLab

[官网](https://dev101.swanlab.cn) · [Github](https://github.com/swanhubx/swanlab) · [贡献指南](https://github.com/SwanHubX/SwanLab/blob/main/CONTRIBUTING.md) · [SwanHub开源社区](https://swanhub.co) · [VSCode插件](https://marketplace.visualstudio.com/items?itemName=SwanLab.swanlab&ssr=false#overview)
[官网](https://swanlab.cn) · [Github](https://github.com/swanhubx/swanlab) · [VSCode插件](https://marketplace.visualstudio.com/items?itemName=SwanLab.swanlab&ssr=false#overview) · [快速开始](/zh/guide_cloud/general/quick-start.md)

<!-- ![](/assets/swanlab-show.png) -->

![alt text](/assets/product-swanlab-1.png)
<!-- ![alt text](/assets/product-swanlab-1.png) -->

::: warning 👋 我们刚刚上线了密码登录功能
在设置-密码与安全性，完成密码设置后，即可使用密码登录。
:::

SwanLab是一款开源、轻量级的AI实验跟踪工具，提供了一个**跟踪、比较、和协作**实验的平台，旨在加速AI研发团队100倍的研发效率。
![alt text](/assets/product-swanlab-1.png)



SwanLab 是一款**开源、轻量**的 AI 模型训练跟踪与可视化工具，提供了一个**跟踪、记录、比较、和协作实验**的平台。

其提供了友好的API和漂亮的界面，结合了超参数跟踪、指标记录、在线协作、实验链接分享、实时消息通知等功能，让您可以快速跟踪ML实验、可视化过程、分享给同伴。
SwanLab 面向人工智能研究者，设计了友好的Python API 和漂亮的UI界面，并提供**训练可视化、自动日志记录、超参数记录、实验对比、多人协同等功能**。在SwanLab上，研究者能基于直观的可视化图表发现训练问题，对比多个实验找到研究灵感，并通过**在线网页**的分享与基于组织的**多人协同训练**，打破团队沟通的壁垒，提高组织训练效率。

借助SwanLab，科研人员可以沉淀自己的每一次训练经验，与合作者无缝地交流和协作，机器学习工程师可以更快地开发可用于生产的模型。



## 为什么使用SwanLab？

与软件工程不同，人工智能是一个**实验性学科**，产生灵感、快速试验、验证想法 是AI研究的主旋律。而记录下实验过程和灵感，就像化学家记录实验手稿一样，是每一个AI研究者、研究组织**形成积累、提升加速度**的核心。
Expand All @@ -34,25 +40,29 @@ SwanLab是一款开源、轻量级的AI实验跟踪工具，提供了一个**跟

## SwanLab能做什么？

![alt text](/assets/how-to-do-swanlab.png)
**1. 📊 实验指标与超参数跟踪**: 极简的代码嵌入您的机器学习 pipeline，跟踪记录训练关键指标

- 支持**云端**使用（类似Weights & Biases），随时随地查看训练进展。[手机看实验的方法](https://docs.swanlab.cn/guide_cloud/general/app.html)
- 支持**超参数记录**与表格展示
- **支持的元数据类型**：标量指标、图像、音频、文本、...
- **支持的图表类型**：折线图、媒体图（图像、音频、文本）、...
- **后台自动记录**：日志logging、硬件环境、Git 仓库、Python 环境、Python 库列表、项目运行目录

**2. ⚡️ 全面的框架集成**: PyTorch、🤗HuggingFace Transformers、PyTorch Lightning、🦙LLaMA Factory、MMDetection、Ultralytics、PaddleDetetion、LightGBM、XGBoost、Keras、Tensorboard、Weights&Biases、OpenAI、Swift、XTuner、Stable Baseline3、Hydra 在内的 **30+** 框架

**1. 📊实验指标与超参数跟踪**: 极简的代码嵌入您的机器学习pipeline，跟踪记录训练关键指标
- 自由的超参数与实验配置记录
- **支持的元数据类型**：标量指标、图像、音频、文本、...
- **支持的图表类型**：折线图、媒体图（图像、音频、文本）、...
- **自动记录**：控制台logging、GPU硬件、Git信息、Python解释器、Python库列表、代码目录
![](/assets/integrations.png)

**2. ⚡️全面的框架集成**: [PyTorch](/zh/guide_cloud/integration/integration-pytorch.md)、[PyTorch Lightning](/zh/guide_cloud/integration/integration-pytorch-lightning.md)、[🤗HuggingFace Transformers](/zh/guide_cloud/integration/integration-huggingface-transformers.md)、[MMEngine](/zh/guide_cloud/integration/integration-mmengine.md)、[Ultralytics](/zh/guide_cloud/integration/integration-ultralytics.md)等主流框架
**3. 💻 硬件监控**: 支持实时记录与监控CPU、NPU（昇腾Ascend）、GPU（英伟达Nvidia）、内存的系统级硬件指标

**3. 📦组织实验**: 集中式仪表板，快速管理多个项目与实验，通过整体视图速览训练全局
**4. 📦 实验管理**: 通过专为训练场景设计的集中式仪表板，通过整体视图速览全局，快速管理多个项目与实验

**4. 🆚比较结果**: 通过在线表格与对比图表比较不同实验的超参数和结果，挖掘迭代灵感
**4. 🆚 比较结果**: 通过在线表格与对比图表比较不同实验的超参数和结果，挖掘迭代灵感

**5. 👥在线协作**: 您可以与团队进行协作式训练，支持将实验实时同步在一个项目下，您可以在线查看团队的训练记录，基于结果发表看法与建议
**5. 👥 在线协作**: 您可以与团队进行协作式训练，支持将实验实时同步在一个项目下，您可以在线查看团队的训练记录，基于结果发表看法与建议

**6. ✉️分享结果**: 复制和发送持久的URL来共享每个实验，方便地发送给伙伴，或嵌入到在线笔记中
**6. ✉️ 分享结果**: 复制和发送持久的 URL 来共享每个实验，方便地发送给伙伴，或嵌入到在线笔记中

**7. 💻支持自托管**: 支持不联网使用，自托管的社区版同样可以查看仪表盘与管理实验
**7. 💻 支持自托管**: 支持离线环境使用，自托管的社区版同样可以查看仪表盘与管理实验


## 从哪里开始
Expand Down

0 comments on commit 24f032b

Please sign in to comment.