@liuhange/dsh-ai-dataset-inspector 3.1.0 → 3.1.1
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +17 -1
- package/lib/__tests__/aiDataset.test.js +18 -0
- package/package.json +5 -3
package/README.md
CHANGED
|
@@ -1,8 +1,17 @@
|
|
|
1
1
|
# @liuhange/dsh-ai-dataset-inspector
|
|
2
2
|
|
|
3
|
+
## 基本信息
|
|
4
|
+
|
|
5
|
+
| 项目 | 值 |
|
|
6
|
+
|------|-----|
|
|
7
|
+
| 插件名 | dsh-ai-dataset-inspector |
|
|
8
|
+
| 包名 | @liuhange/dsh-ai-dataset-inspector |
|
|
9
|
+
| 版本 | 3.1.0 |
|
|
10
|
+
| 工具 | inspect_ai_dataset |
|
|
11
|
+
|
|
3
12
|
## 功能说明
|
|
4
13
|
|
|
5
|
-
AI
|
|
14
|
+
检查AI训练数据集质量，涵盖标注一致性（Kappa系数）、标签均衡性（类别分布比）、数据泄露检测（训练/验证集重叠）等维度。支持引用data-masking结果检测PII残留。
|
|
6
15
|
|
|
7
16
|
## 安装
|
|
8
17
|
|
|
@@ -17,6 +26,13 @@ import { apply } from '@liuhange/dsh-ai-dataset-inspector';
|
|
|
17
26
|
apply(ctx); // 注册工具: inspect_ai_dataset
|
|
18
27
|
```
|
|
19
28
|
|
|
29
|
+
## 政策依据
|
|
30
|
+
|
|
31
|
+
| 文件名称 | 文号 | 核心条款摘要 |
|
|
32
|
+
|---------|------|------------|
|
|
33
|
+
| 关于推进行业高质量数据集建设行动的实施方案 | 国数发〔2026〕39号 | 部署强基扩容、标注攻坚、提质增效、应用赋能、管理服务、价值释放六大行动 |
|
|
34
|
+
| 高质量数据集建设指南 | T/TC609 005-2025 | 覆盖数据需求、规划、采集、预处理、标注、模型验证全生命周期 |
|
|
35
|
+
|
|
20
36
|
## License
|
|
21
37
|
|
|
22
38
|
MIT
|
|
@@ -0,0 +1,18 @@
|
|
|
1
|
+
import { describe, it, expect } from 'vitest';
|
|
2
|
+
describe('ai-dataset-inspector', () => {
|
|
3
|
+
it('导出name/inject/apply', async () => {
|
|
4
|
+
const mod = await import('../index.js');
|
|
5
|
+
expect(mod.name).toBe('@liuhange/dsh-ai-dataset-inspector');
|
|
6
|
+
expect(mod.inject).toEqual(['tools']);
|
|
7
|
+
expect(typeof mod.apply).toBe('function');
|
|
8
|
+
});
|
|
9
|
+
it('apply注册inspect_ai_dataset工具', async () => {
|
|
10
|
+
const mod = await import('../index.js');
|
|
11
|
+
const tools = [];
|
|
12
|
+
mod.apply({ tools: { register: (t) => tools.push(t) } });
|
|
13
|
+
const tool = tools[0];
|
|
14
|
+
expect(tool.name).toBe('inspect_ai_dataset');
|
|
15
|
+
expect(tool.description).toContain('AI');
|
|
16
|
+
});
|
|
17
|
+
});
|
|
18
|
+
//# sourceMappingURL=aiDataset.test.js.map
|
package/package.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "@liuhange/dsh-ai-dataset-inspector",
|
|
3
3
|
"description": "AI training dataset inspection plugin: annotation consistency, label distribution, data leakage detection",
|
|
4
|
-
"version": "3.1.
|
|
4
|
+
"version": "3.1.1",
|
|
5
5
|
"publishConfig": {
|
|
6
6
|
"access": "public"
|
|
7
7
|
},
|
|
@@ -31,10 +31,12 @@
|
|
|
31
31
|
"test": "vitest run",
|
|
32
32
|
"build": "tsc"
|
|
33
33
|
},
|
|
34
|
+
"dependencies": {
|
|
35
|
+
"@liuhange/dsh-data-asset-shared": "^3.1.0"
|
|
36
|
+
},
|
|
34
37
|
"peerDependencies": {
|
|
35
38
|
"@deepseek-ai/cordis": "^4.0.0",
|
|
36
|
-
"@deepseek-ai/dsh-tools": "0.0.1-rc.1"
|
|
37
|
-
"@liuhange/dsh-data-asset-shared": "^3.0.0"
|
|
39
|
+
"@deepseek-ai/dsh-tools": "0.0.1-rc.1"
|
|
38
40
|
},
|
|
39
41
|
"dsh": {
|
|
40
42
|
"bundle": {
|