zywen commited on
Commit
f837e77
·
verified ·
1 Parent(s): f7da51b

design: simplify organization card layout

Browse files
Files changed (3) hide show
  1. README.md +15 -41
  2. README.zh-CN.md +16 -42
  3. assets/hero-edge-ai.webp +0 -3
README.md CHANGED
@@ -8,55 +8,29 @@ pinned: false
8
  ---
9
 
10
  <p align="right">
11
- <strong>EN</strong> ·
12
  <a href="https://huggingface.co/spaces/Hansoric/README/blob/main/README.zh-CN.md">简体中文</a>
13
  </p>
14
 
15
- <p align="center">
16
- <img src="https://huggingface.co/spaces/Hansoric/README/resolve/main/assets/hero-edge-ai.webp" alt="Distributed edge AI systems connected through a coordinated network" width="100%" />
17
- </p>
18
-
19
- <h1 align="center">Hansoric</h1>
20
 
21
- <p align="center"><strong>Engineering AI agent systems for real-world edge deployment.</strong></p>
22
 
23
- <p align="center">
24
- <code>EDGE AI</code>&nbsp;&nbsp;
25
- <code>AGENT SYSTEMS</code>&nbsp;&nbsp;
26
- <code>OPEN TECHNOLOGY</code>
27
- </p>
28
-
29
- <p align="center">
30
- We are building practical components for AI agents that operate beyond the cloud.<br />
31
- This organization is the home for our public models, datasets, evaluations, and developer tools.
32
- </p>
33
-
34
- <p align="center">
35
- <a href="https://hansoric.com/en"><strong>Visit hansoric.com →</strong></a>&nbsp;&nbsp;·&nbsp;&nbsp;
36
- <a href="https://huggingface.co/Hansoric"><strong>Follow on Hugging Face</strong></a>
37
- </p>
38
-
39
- ---
40
 
41
- ## Engineered for the edge
42
 
43
- Edge agents work under real limits. Our public work keeps those constraints visible from design through evaluation.
 
 
 
44
 
45
- - **⚡ Resource-aware** — Compute, memory, latency, and connectivity are first-class constraints.
46
- - **◎ Verifiable** — Expected behavior, limitations, and failure modes are evaluated explicitly.
47
- - **↻ Reproducible** — Configurations, versions, and evaluation conditions are documented clearly.
48
 
49
- ## Open release areas
50
 
51
- <table>
52
- <tr>
53
- <td width="50%"><strong>◈ Edge Models</strong><br />Efficient models and reusable components for constrained environments.</td>
54
- <td width="50%"><strong>▦ Datasets</strong><br />Structured resources for developing and validating edge-agent tasks.</td>
55
- </tr>
56
- <tr>
57
- <td width="50%"><strong>◉ Evals</strong><br />Repeatable evaluations for capability, reliability, and efficiency.</td>
58
- <td width="50%"><strong>⌘ Tools</strong><br />Utilities for experimentation, integration, and edge deployment.</td>
59
- </tr>
60
- </table>
61
 
62
- > **Release status:** Our first public artifacts are in preparation. Direct links will appear here as documented resources become available.
 
8
  ---
9
 
10
  <p align="right">
11
+ <strong>English</strong> ·
12
  <a href="https://huggingface.co/spaces/Hansoric/README/blob/main/README.zh-CN.md">简体中文</a>
13
  </p>
14
 
15
+ # AI agent systems, engineered for the edge.
 
 
 
 
16
 
17
+ Hansoric builds practical components for AI agents that operate beyond the cloud. This organization is where we publish the models, datasets, evaluations, and tools behind that work.
18
 
19
+ [Visit hansoric.com →](https://hansoric.com/en)
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
20
 
21
+ ## What we publish
22
 
23
+ - **Edge models** — Efficient models and reusable components for constrained environments.
24
+ - **Datasets** — Structured resources for developing and validating edge-agent tasks.
25
+ - **Evaluations** — Repeatable evaluations for capability, reliability, and efficiency.
26
+ - **Developer tools** — Utilities for experimentation, integration, and edge deployment.
27
 
28
+ ## Engineering principles
 
 
29
 
30
+ Edge agents work under real constraints. We keep those limits visible from design through evaluation.
31
 
32
+ - **Resource-aware** — Compute, memory, latency, and connectivity are first-class constraints.
33
+ - **Verifiable** — Expected behavior, limitations, and failure modes are evaluated explicitly.
34
+ - **Reproducible** — Configurations, versions, and evaluation conditions are documented clearly.
 
 
 
 
 
 
 
35
 
36
+ > Our first public releases are in preparation. Direct links will be added as documented resources become available.
README.zh-CN.md CHANGED
@@ -1,53 +1,27 @@
1
  <p align="right">
2
- <a href="https://huggingface.co/Hansoric">EN</a> ·
3
- <strong>中文</strong>
4
  </p>
5
 
6
- <p align="center">
7
- <img src="https://huggingface.co/spaces/Hansoric/README/resolve/main/assets/hero-edge-ai.webp" alt="通过协同网络连接的分布式边缘智能系统" width="100%" />
8
- </p>
9
-
10
- <h1 align="center">Hansoric</h1>
11
-
12
- <p align="center"><strong>面向真实场景边缘部署的 AI 智能体系统工程。</strong></p>
13
-
14
- <p align="center">
15
- <code>边缘智能</code>&nbsp;&nbsp;
16
- <code>智能体系统</code>&nbsp;&nbsp;
17
- <code>开放技术</code>
18
- </p>
19
-
20
- <p align="center">
21
- 我们正在构建让 AI 智能体走出云端的实用技术组件。<br />
22
- 本组织将用于发布公开模型、数据集、评测资源与开发工具。
23
- </p>
24
 
25
- <p align="center">
26
- <a href="https://hansoric.com/zh"><strong>访问 hansoric.com →</strong></a>&nbsp;&nbsp;·&nbsp;&nbsp;
27
- <a href="https://huggingface.co/Hansoric"><strong>关注 Hugging Face</strong></a>
28
- </p>
29
 
30
- ---
31
 
32
- ## 为边缘而工程化
33
 
34
- 边缘智能体始终面对真实资源边界。我们的公开技术工作会让这些约束贯穿设计与评测全过程。
 
 
 
35
 
36
- - **⚡ 资源感知** — 将算力、内存、时延与网络连接视为一级约束。
37
- - **◎ 可验证** — 明确评估预期行为、适用边界与失败模式。
38
- - **↻ 可复现** — 清晰记录配置、版本与评测条件。
39
 
40
- ## 开放发布方向
41
 
42
- <table>
43
- <tr>
44
- <td width="50%"><strong>◈ 边缘模型</strong><br />面向资源受限环境的高效模型与可复用组件。</td>
45
- <td width="50%"><strong>▦ 数据集</strong><br />支撑边缘智能体任务开发与验证的结构化资源。</td>
46
- </tr>
47
- <tr>
48
- <td width="50%"><strong>◉ 评测</strong><br />用于衡量能力、可靠性与效率的可重复评测。</td>
49
- <td width="50%"><strong>⌘ 工具</strong><br />支持实验、集成与边缘部署的开发工具。</td>
50
- </tr>
51
- </table>
52
 
53
- > **发布状态:** 首批公开资源正在准备中。相关内容完成文档化并达到公开标准后,本页将补充直接入口。
 
1
  <p align="right">
2
+ <a href="https://huggingface.co/Hansoric">English</a> ·
3
+ <strong>简体中文</strong>
4
  </p>
5
 
6
+ # 面向边缘部署的 AI 智能体系统工程
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
7
 
8
+ Hansoric 构建让 AI 智能体走出云端、运行于真实场景的实用技术组件。本组织用于发布支撑这些工作的模型、数据集、评测资源与开发工具。
 
 
 
9
 
10
+ [访问 hansoric.com →](https://hansoric.com/zh)
11
 
12
+ ## 我们发布什么
13
 
14
+ - **边缘模型** — 面向资源受限环境的高效模型与可复用组件。
15
+ - **数据集** — 支撑边缘智能体任务开发与验证的结构化资源。
16
+ - **评测** — 用于衡量能力、可靠性与效率的可重复评测。
17
+ - **开发工具** — 支持实验、集成与边缘部署的实用工具。
18
 
19
+ ## 工程原则
 
 
20
 
21
+ 边缘智能体始终面对真实资源边界。我们让这些约束贯穿设计与评测全过程。
22
 
23
+ - **资源感知** — 将算力、内存、时延与网络连接视为一级约束。
24
+ - **可验证** — 明确评估预期行为、适用边界与失败模式。
25
+ - **可复现** — 清晰记录配置、版本与评测条件。
 
 
 
 
 
 
 
26
 
27
+ > 首批公开资源正在准备中。相关内容完成文档化并达到公开标准后,本页将补充直接入口。
assets/hero-edge-ai.webp DELETED

Git LFS Details

  • SHA256: ebde758b7ca0373313e404400c4e7a364af2dde676959a5398e9f43a0353117d
  • Pointer size: 131 Bytes
  • Size of remote file: 189 kB