Skip to content

Commit 8fc10ec

Browse files
ditaccms-botditaccms-bot
committed
Update from SAP DITA CMS (squashed):
commit a6819a351031e0fbce085dfa2a316c4a9ef39783 Author: REDACTED Date: Mon Nov 3 17:53:24 2025 +0000 Update from SAP DITA CMS 2025-11-03 17:53:24 Project: dita-all/ayn1620809929290 Project map: 8a3894192639433b9ba1f87efe78bfd1.ditamap Output: loio62559d01add8414eb0c4d76d2d9f48bd Language: en-US Builddable map: 038a6194f65c4ef68885f6f16360dbc4.ditamap commit 9e32f45544f044ffda45eec80286cc7511147cec Author: REDACTED Date: Mon Nov 3 17:38:21 2025 +0000 Update from SAP DITA CMS 2025-11-03 17:38:21 Project: dita-all/ayn1620809929290 Project map: 8a3894192639433b9ba1f87efe78bfd1.ditamap Output: loio62559d01add8414eb0c4d76d2d9f48bd Language: en-US Builddable map: 038a6194f65c4ef68885f6f16360dbc4.ditamap commit 88a379c49c5070369c8c730d7e2b6f3c4e124f2a Author: REDACTED Date: Mon Nov 3 16:10:39 2025 +0000 Update from SAP DITA CMS 2025-11-03 16:10:39 Project: dita-all/bex1621329160251 Project map: d3e749bbac3d4f728c12228db6629c45.ditamap ################################################## [Remaining squash message was removed before commit...]
1 parent 5fb8774 commit 8fc10ec

File tree

56 files changed

+3937
-1369
lines changed

Some content is hidden

Large Commits have some content hidden by default. Use the searchbox below for content that may be hidden.

56 files changed

+3937
-1369
lines changed

‎docs/sap-ai-core/content-filtering-f804175.md‎

Lines changed: 10 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -36,6 +36,16 @@ A prompt attack is a malicious input that is designed to bypass a model's safety
3636

3737
If both prompt attack detection and harm classification are configured, the orchestration service performs the prompt attack detection call first, and then performs the harm classification call. If a prompt attack is detected, the orchestration service does not make the harm classification request and returns the prompt attack detection result only.
3838

39+
40+
41+
### Protected Material Detection for Code
42+
43+
The Azure Content Safety service includes an output content filter that identifies code matching existing GitHub repository code.
44+
45+
The filter scans publicly available code, and has a knowledge cutoff date. Only code matching publicly available code added to GitHub after the knowledge cutoff date can be detected.
46+
47+
For more information see [Protected Material Detection for Code in Azure AI Content Safety](https://learn.microsoft.com/en-us/azure/ai-services/content-safety/quickstart-protected-material-code?pivots=programming-language-rest).
48+
3949
> ### Remember:
4050
> When multiple content filtering types are configured together, each type is processed as a separate request. Costs will be incurred for each filtering type applied.
4151

‎docs/sap-ai-core/create-a-configuration-for-an-optimization-40ba168.md‎

Lines changed: 258 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,258 @@
1+
<!-- loio40ba1683832d4c3a92eaac8c0c399b18 -->
2+
3+
# Create a Configuration for an Optimization
4+
5+
6+
7+
## Prerequisites
8+
9+
- You're using the `extended` service plan. For more information, see [Service Plans](service-plans-c7244c6.md).
10+
- You've registered an object store secret with the name `default` for output artifacts. For more information, see [Register an Object Store for Optimizations](register-an-object-store-for-optimizations-54068a9.md).
11+
- If you want to separate your input and output object stores, you've registered an object store secret for input artifacts with a name of your choice. For more information, see [Register an Object Store for Optimizations](register-an-object-store-for-optimizations-54068a9.md).
12+
- You've prepared an optimization dataset and registered it as an optimization artifact. For more information, see [Dataset preparation](dataset-preparation-b2625d7.md) and [Register Optimization Artifacts](register-optimization-artifacts-b8a9cd8.md).
13+
14+
- You've prepared a prompt template, and your template is available in the prompt registry. For more information, see [Template preparation](template-preparation-4526dde.md).
15+
16+
17+
18+
19+
## Procedure
20+
21+
Create a configuration by sending a POST request to endpoint `$AI_API_URL/v2/lm/configurations`.
22+
23+
Include the following in your request:
24+
25+
26+
<table>
27+
<tr>
28+
<th valign="top">
29+
30+
Field
31+
32+
</th>
33+
<th valign="top" colspan="2">
34+
35+
Value
36+
37+
</th>
38+
</tr>
39+
<tr>
40+
<td valign="top">
41+
42+
$AI\_API\_URL
43+
44+
</td>
45+
<td valign="top" colspan="2">
46+
47+
The base URL of your SAP AI Core environment. This can also be set as an environment variable.
48+
49+
</td>
50+
</tr>
51+
<tr>
52+
<td valign="top">
53+
54+
$RESOURCE\_GROUP
55+
56+
</td>
57+
<td valign="top" colspan="2">
58+
59+
The AI resource group assigned to your account
60+
61+
</td>
62+
</tr>
63+
<tr>
64+
<td valign="top">
65+
66+
$TOKEN
67+
68+
</td>
69+
<td valign="top" colspan="2">
70+
71+
Your access token for SAP AI Core
72+
73+
</td>
74+
</tr>
75+
<tr>
76+
<td valign="top">
77+
78+
`inputArtifactBindings`
79+
80+
</td>
81+
<td valign="top" colspan="2">
82+
83+
The `artifactId` for your registered dataset
84+
85+
</td>
86+
</tr>
87+
<tr>
88+
<td valign="top">
89+
90+
`scenarioId`
91+
92+
</td>
93+
<td valign="top" colspan="2">
94+
95+
`genai-optimizations`
96+
97+
</td>
98+
</tr>
99+
<tr>
100+
<td valign="top">
101+
102+
`executableId`
103+
104+
</td>
105+
<td valign="top" colspan="2">
106+
107+
`genai-optimizations`
108+
109+
</td>
110+
</tr>
111+
<tr>
112+
<td valign="top" rowspan="7">
113+
114+
`parameterBindings`
115+
116+
</td>
117+
<td valign="top">
118+
119+
`basePrompt` \(mandatory\)
120+
121+
</td>
122+
<td valign="top">
123+
124+
Reference to the prompt in the prompt registry that should be optimized in the format “`<scenario>/<promptName>:<promptVersion>`”
125+
126+
</td>
127+
</tr>
128+
<tr>
129+
<td valign="top">
130+
131+
`baseModel`\(optional\)
132+
133+
</td>
134+
<td valign="top">
135+
136+
In case an evaluation of the selected metric should be run against a baseline model without optimization, to determine the pre-optimization metric. Default: `none`.
137+
138+
</td>
139+
</tr>
140+
<tr>
141+
<td valign="top">
142+
143+
`dataset` \(mandatory\)
144+
145+
</td>
146+
<td valign="top">
147+
148+
The dataset that is used as basis for optimization. The dataset contains desired responses from the target model. The optimization will maximize the optimization metric by varying the input prompt. This parameter specifies the path to the json contained in the input artifact that is submitted as part of the configuration. The dataset must contain a minimum of 25 samples and a maximum of 200 samples.
149+
150+
</td>
151+
</tr>
152+
<tr>
153+
<td valign="top">
154+
155+
`optimizationMetric` \(mandatory\)
156+
157+
</td>
158+
<td valign="top">
159+
160+
The metric to be used for optimizing. The possible choices are:
161+
162+
- `LLMaaJ:Sem_Sim_10`
163+
164+
This metric uses an LLM-as-a-judge to evaluate the semantic similarity of the model response relative target dataset on a scale of 1-10.
165+
166+
167+
- `JSON_Match`
168+
169+
This metric determines if the LLM's JSON output matches the dataset output. The precision, recall, and f1 scores are computed based on individual fields in the JSON and averaged across all samples in the dataset. This is useful for applications that require structured outputs.
170+
171+
172+
173+
174+
</td>
175+
</tr>
176+
<tr>
177+
<td valign="top">
178+
179+
`targetModels` \(mandatory\)
180+
181+
</td>
182+
<td valign="top">
183+
184+
A comma separated list of models with version for which to optimize the prompt. A prompt can be optimized for a maximum of four models. The format is `<modelName>:<modelVersion>`. The target models must be available in the corresponding region. For more information, see SAP Note [3437766](https://me.sap.com/notes/3437766).
185+
186+
</td>
187+
</tr>
188+
<tr>
189+
<td valign="top">
190+
191+
`targetPromptMapping` \(mandatory\)
192+
193+
</td>
194+
<td valign="top">
195+
196+
Provides a mapping to the optimized prompt for storage in the prompt registry in the format `<modelName>:<modelVersion>=<promptName>:<promptVersion>`.
197+
198+
</td>
199+
</tr>
200+
<tr>
201+
<td valign="top">
202+
203+
`includeFewShotExamples` \(optional\)
204+
205+
</td>
206+
<td valign="top">
207+
208+
Whether to include some parts of the provided dataset in the input prompt to improve the output. Default: `false`.
209+
210+
</td>
211+
</tr>
212+
</table>
213+
214+
For example:
215+
216+
> ### Sample Code:
217+
> ```
218+
> curl --request POST "$AI_API_URL/v2/lm/configurations" \
219+
> --header "Authorization: Bearer $TOKEN" \
220+
> --header "AI-Resource-Group: $RESOURCE_GROUP" \
221+
> --header "Content-Type: application/json" \
222+
> --data '{
223+
> "name": "genai-eval-conf",
224+
> "scenarioId": "genai-optimizations ",
225+
> "executableId": "genai-optimizations ",
226+
> "inputArtifactBindings": [
227+
> {
228+
> "key": "datasetFolder",
229+
> "artifactId": "1f50fbf8-18d6-4e72-ae10-04fdbc087815"
230+
> }
231+
> ],
232+
> "parameterBindings": [
233+
> {
234+
> "key": "basePrompt",
235+
> "value": "<scenario>/<promptName>:<promptVersion>"
236+
> },
237+
> {
238+
> "key": "dataset",
239+
> "value": "data.json",
240+
> },
241+
> {
242+
> "key": "optimizationMetric",
243+
> "value": "JSON_Match”
244+
> },
245+
> {
246+
>
247+
> "key": "targetModels",
248+
> "value": "<modelName>:<modelVersion>”
249+
> },
250+
> {
251+
>
252+
> "key": "targetPromptMapping",
253+
> "value": “<modelName>:<modelversion>=<promptName>:<promptVersion>”
254+
> },
255+
> ]
256+
> }‘
257+
> ```
258+

‎docs/sap-ai-core/create-a-document-grounding-pipeline-using-the-pipelines-api-0a13e1c.md‎

Lines changed: 65 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,65 @@
1+
<!-- copy0a13e1cf4cbf4686ac16274c48f19d1f -->
2+
3+
# Create a Document Grounding Pipeline Using the Pipelines API
4+
5+
This API call creates a pipeline for indexing documents for a resource group.
6+
7+
8+
9+
<a name="copy0a13e1cf4cbf4686ac16274c48f19d1f__section_ljc_ksf_ngc"/>
10+
11+
## Prerequisites
12+
13+
- You have created a resource group. For more information, see [Create a Resource Group for Grounding](create-a-resource-group-for-grounding-e32efa5.md).
14+
- You have created a generic secret. For more information, see [Generic Secrets for Grounding](generic-secrets-for-grounding-e1a201c.md).
15+
- You've added your grounding documents to your document store and you're using a supported data source and document type. For more information, see [Pipelines API](pipelines-api-d8cc0e3.md).
16+
- If you're using metadata, you've configured a metadata server. For more information, see [Prepare your Metadata API Server](prepare-your-metadata-api-server-23a0741.md).
17+
18+
19+
20+
<a name="copy0a13e1cf4cbf4686ac16274c48f19d1f__section_d3p_kfv_lfc"/>
21+
22+
## Context
23+
24+
You can schedule automatic content updates by using a `cronExpression` when you create your pipeline. For more information, see [Cron Expressions](cron-expressions-6175008.md).
25+
26+
> ### Tip:
27+
> If you use the pipelines API, you do not need to call the Vector API separately. After the data is embedded, you can directly use the Retrieval API to query the vector store for relevant sections.
28+
29+
30+
31+
<a name="copy0a13e1cf4cbf4686ac16274c48f19d1f__section_sbw_lw2_pgc"/>
32+
33+
## Procedure
34+
35+
You document grounding pipeline with your chosen repository. For more information, see the following:
36+
37+
- [Create a Pipeline with Microsoft SharePoint](create-a-pipeline-with-microsoft-sharepoint-80326fe.md)
38+
- [Create a Pipeline with AWS S3](create-a-pipeline-with-aws-s3-7f97adf.md)
39+
- [Create a Pipeline with SFTP](create-a-pipeline-with-sftp-3d95d74.md)
40+
- [Create a Pipeline with SAP Build Work Zone](create-a-pipeline-with-sap-build-work-zone-d499612.md)
41+
- [Create a Pipeline with SAP Document Management Service](create-a-pipeline-with-sap-document-management-service-5718584.md)
42+
43+
44+
45+
<a name="copy0a13e1cf4cbf4686ac16274c48f19d1f__section_tgp_1x2_pgc"/>
46+
47+
## Next Steps
48+
49+
You can use API endpoints to view and manage the lifecycle of your data pipelines, executions and documents. For more information, see [Data Pipelines](data-pipelines-9d9eb37.md), [Executions](executions-c70b70d.md) and [Documents](documents-f87ab02.md).
50+
51+
You can manually restart a pipeline. For more information, see [Manually Restart a Document Grounding Pipeline](manually-restart-a-document-grounding-pipeline-932f2b7.md).
52+
53+
- **[Create a Pipeline with Microsoft SharePoint](create-a-pipeline-with-microsoft-sharepoint-80326fe.md "")**
54+
55+
- **[Create a Pipeline with AWS S3](create-a-pipeline-with-aws-s3-7f97adf.md "")**
56+
57+
- **[Create a Pipeline with SFTP](create-a-pipeline-with-sftp-3d95d74.md "")**
58+
59+
- **[Create a Pipeline with SAP Build Work Zone](create-a-pipeline-with-sap-build-work-zone-d499612.md "")**
60+
61+
- **[Create a Pipeline with SAP Document Management Service](create-a-pipeline-with-sap-document-management-service-5718584.md "")**
62+
63+
- **[Cron Expressions](cron-expressions-515e839.md "")**
64+
65+

0 commit comments

Comments
 (0)