Skip to content

Commit 3eb6467

Browse files
ditaccms-botditaccms-bot
committed
Update from SAP DITA CMS (squashed):
commit 4ea0c3ca2321911f67bdd78386fc3df222acdb7c Author: REDACTED Date: Tue Apr 29 09:16:08 2025 +0000 Update from SAP DITA CMS 2025-04-29 09:16:08 Project: dita-all/ayn1620809929290 Project map: 8a3894192639433b9ba1f87efe78bfd1.ditamap Output: loio62559d01add8414eb0c4d76d2d9f48bd Language: en-US Builddable map: 038a6194f65c4ef68885f6f16360dbc4.ditamap commit bdf84a0378c8e5703cc7d6859c2ff717db2c7835 Author: REDACTED Date: Tue Apr 29 09:13:32 2025 +0000 Update from SAP DITA CMS 2025-04-29 09:13:32 Project: dita-all/bex1621329160251 Project map: d3e749bbac3d4f728c12228db6629c45.ditamap Output: loiodaa66b2ef49f48539fa2882d82d5b619 Language: en-US Builddable map: f17fa8568d0448c685f2a0301061a6ee.ditamap commit d96d06648bab248862b0862220c6370a53840ba8 Author: REDACTED Date: Mon Apr 28 08:54:50 2025 +0000 Update from SAP DITA CMS 2025-04-28 08:54:50 Project: dita-all/bex1621329160251 Project map: d3e749bbac3d4f728c12228db6629c45.ditamap ################################################## [Remaining squash message was removed before commit...]
1 parent 7369bb4 commit 3eb6467

File tree

51 files changed

+753
-77
lines changed

Some content is hidden

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

51 files changed

+753
-77
lines changed

‎docs/sap-ai-core/accessing-generative-ai-models-through-global-scenarios-4ca11f6.md‎

Lines changed: 3 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -16,19 +16,17 @@ Access to generative AI models falls under the global AI scenarios `foundation-m
1616

1717
- Provider Agnostic: The orchestration service offers a unified interface for accessing LLMs from multiple providers, simplifying the process of testing and comparing models. You can switch models via a config parameter without creating new deployments.
1818
- Simplified Integration: It streamlines the integration of LLMs into applications, ensuring efficient and cost-effective development.
19-
- Expandability: The service allows you to start with basic features and add modules as needed, offering a smooth learning curve.
19+
- Expandability: The service allows you to start with basic features and add modules such as content filtering, anonymization, or grounding as needed, offering a smooth learning curve.
2020

2121

2222

2323
### Considerations Before Using Orchestration
2424

25-
- Function Calling Availability: Function calling, with its structured output aspects, isn't currently available.
2625
- Open Source Framework Support: While we support LangChain, a popular open-source framework, we can't support every framework available. If a specific open-source framework is required, you might need to extend it for support or consider alternative integration methods.
2726

2827
- **[Orchestration](orchestration-cdd4847.md "The orchestration service operates under the global AI scenario
29-
orchestration, which is managed by SAP AI Core. This service
30-
enables the use of various generative AI models with a unified code, configuration, and
31-
deployment.")**
28+
orchestration, which is managed by SAP AI Core. This service enables the use of various generative AI
29+
models with a unified code, configuration, and deployment.")**
3230
The orchestration service operates under the global AI scenario `orchestration`, which is managed by SAP AI Core. This service enables the use of various generative AI models with a unified code, configuration, and deployment.
3331
- **[Foundation Models](foundation-models-2d981fb.md "The foundation models service operates under the global AI scenario
3432
foundation-models, which is managed by SAP AI Core.")**

‎docs/sap-ai-core/advanced-features-24f2fbb.md‎

Lines changed: 3 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -5,8 +5,9 @@
55
Explore advanced features, within SAP AI Core.
66

77
- **[AI Content as a Service](ai-content-as-a-service-3fb0390.md "SAP AI Core helps users provide their AI content available as a
8-
service on the Service Marketplace, using GitOps.")**
9-
SAP AI Core helps users provide their AI content available as a service on the *Service Marketplace*, using GitOps.
8+
service on the SAP BTP
9+
Service Marketplace, using GitOps.")**
10+
SAP AI Core helps users provide their AI content available as a service on the SAP BTP *Service Marketplace*, using GitOps.
1011

1112
**Parent topic:**[Predictive AI](predictive-ai-6c3b730.md "")
1213

‎docs/sap-ai-core/ai-content-as-a-service-3fb0390.md‎

Lines changed: 29 additions & 11 deletions
Original file line numberDiff line numberDiff line change
@@ -2,35 +2,53 @@
22

33
# AI Content as a Service
44

5-
SAP AI Core helps users provide their AI content available as a service on the *Service Marketplace*, using GitOps.
5+
SAP AI Core helps users provide their AI content available as a service on the SAP BTP *Service Marketplace*, using GitOps.
66

7-
Consumers can create service instances using SAP AI Core to use available content, and create their own executions or deployments. If the person acting as the main tenant chooses to make this available externally through the *Service Marketplace*, they become a service provider. The onboarding process for service providers and their content is as follows:
7+
**Service Providers** are an entity acting as SAP AI Core main tenant, who have chosen to make their AI Content \(for example Serving Template, Workflow Template, Docker Container\) available for consumption as a service through SAP BTP *Service Marketplace*.
8+
9+
**Service Consumer** are an entity acting as a consumer for Service exposing AI content from a Service Provider. Consumers can create service instance and use AI API \(serviceUrl\) to use available AI content. They can start executions or deployments.
10+
11+
12+
13+
<a name="loio3fb039080ce14bad9a773e1a473ce24e__section_wrl_ts3_1fc"/>
14+
15+
## Service Provider Flow
816

917
1. Create AI content such as a workflow, serving template, or Docker registry that is consumer ready.
1018
2. Create a generic secret for broker registration. For more information, see [Create a Generic Secret](create-a-generic-secret-1831845.md).
1119
3. Provide a service custom resource YAML in a registered git repository.
12-
4. SAP AI Core creates a service broker for the content to be made available to the *Service Marketplace*. The service broker handles onboarding and offboarding for end consumers.
20+
4. SAP AI Core creates a service broker for the content to be made available to the SAP BTP *Service Marketplace*. The service broker handles onboarding and offboarding for end consumers.
1321
5. Fetch the service broker information by calling the endpoint: `{{apiurl}}/v2/admin/services`.
14-
6. Register the service broker in the *Service Marketplace* and SAP Cloud Management service.
15-
7. The consumer creates service instance from the *Service Marketplace*.
16-
8. SAP AI Core creates a resource group for the consumer with *<resourceGroupId\>* == `serviceInstanceId`.
17-
9. The consumer creates a service key and starts using the service.
22+
6. Register the service broker in the SAP BTP *Service Marketplace* and SAP Cloud Management service.
23+
24+
![](images/SAAS_producer_4a_ecde38a.png)
25+
26+
27+
28+
<a name="loio3fb039080ce14bad9a773e1a473ce24e__section_g43_ws3_1fc"/>
29+
30+
## Service Conusmer Flow
31+
32+
1. The consumer creates service instance from the SAP BTP *Service Marketplace*.
33+
2. SAP AI Core creates a resource group for the consumer with *<resourceGroupId\>* == `serviceInstanceId`.
34+
3. The consumer creates a service key and starts using the service.
1835

1936
> ### Note:
2037
> The service provider main tenant is allowed to provision only 1 Service.
2138
39+
4. Consumer: Starts consuming the service via provided url.
2240

23-
SAP AI Core deploys a generic service broker instance for the main tenant as follows:
24-
25-
![](images/AICoreStructure_31f91a8.png)
41+
![](images/SAAS_consumer_4a_08451cf.png)
2642

2743
- **[Service Custom Resource](service-custom-resource-59f767c.md "The service provider main tenant needs to prepare the service custom resource. The custom resource contains service details, reference to
2844
broker credentials or secrets, and capabilities configured for service consumers.")**
2945
The service provider main tenant needs to prepare the service custom resource. The custom resource contains service details, reference to broker credentials or secrets, and capabilities configured for service consumers.
30-
- **[Onboarding](onboarding-50a6d9f.md "")**
46+
- **[Getting Started as a Service Provider](getting-started-as-a-service-provider-50a6d9f.md "")**
3147

3248
- **[Offboarding](offboarding-1a33323.md "")**
3349

50+
- **[Shared Resource Group](shared-resource-group-dc1e577.md "")**
51+
3452

3553
**Parent topic:**[Advanced Features](advanced-features-24f2fbb.md "Explore advanced features, within SAP AI Core.")
3654

‎docs/sap-ai-core/create-a-deployment-for-a-generative-ai-model-b32e7a8.md‎

Lines changed: 176 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -2,7 +2,7 @@
22

33
# Create a Deployment for a Generative AI Model
44

5-
You make a generative AI model available for use by creating a deployment. You can do so one time for each model and model version, and for each resource group that you want to use with generative AI hub. The deployment URL that is generated can be reused.
5+
66

77

88

@@ -21,6 +21,181 @@ You make a generative AI model available for use by creating a deployment. You c
2121

2222
You make a model available for use by creating a deployment. You can do so one time for each model and model version. The model deployment includes the `modelName` and `version` of the model you want to access. After the deployment is complete, you have a `deploymentUrl`, which can be used across your organization to access the model version.
2323

24+
You can view the available models and their details by sending a GET request to the endpoint `{{apiurl}}/v2/lm/scenarios/foundation-models/models`
25+
26+
```
27+
curl --location '$AI_API_URL/v2/lm/scenarios/foundation-models/models' \
28+
--header 'AI-Resource-Group: default' \
29+
--header 'Authorization: Bearer $AUTH_TOKEN'
30+
31+
```
32+
33+
> ### Output Code:
34+
> ```
35+
> {
36+
> "count": 41,
37+
> "resources": [
38+
> {
39+
> "accessType": "Remote",
40+
> "allowedScenarios": [
41+
> {
42+
> "executableId": "azure-openai",
43+
> "scenarioId": "foundation-models"
44+
> },
45+
> {
46+
> "executableId": "orchestration",
47+
> "scenarioId": "orchestration"
48+
> }
49+
> ],
50+
> "description": "OpenAI GPT-4o engineered for speed and efficiency, supporting text, images and audio using Chat Completions API",
51+
> "displayName": "GPT-4o",
52+
> "executableId": "azure-openai",
53+
> "model": "gpt-4o",
54+
> "provider": "OpenAI",
55+
> "versions": [
56+
> {
57+
> "capabilities": [
58+
> "text-generation",
59+
> "image-recognition"
60+
> ],
61+
> "contextLength": 128000,
62+
> "cost": [
63+
> {
64+
> "inputCost": "0.00312"
65+
> },
66+
> {
67+
> "outputCost": "0.0092"
68+
> }
69+
> ],
70+
> "deprecated": false,
71+
> "inputTypes": [
72+
> "text",
73+
> "image",
74+
> "audio"
75+
> ],
76+
> "isLatest": false,
77+
> "metadata": [
78+
> {
79+
> "meanWinRate": "0.938"
80+
> },
81+
> {
82+
> "chatBotArenaScore": "1285"
83+
> },
84+
> {
85+
> "airBenchRefusalRate": "0.528"
86+
> }
87+
> ],
88+
> "name": "2024-05-13",
89+
> "retirementDate": "",
90+
> "streamingSupported": true
91+
> }, ...
92+
>
93+
> ]
94+
> }
95+
> ```
96+
97+
The output contains the following parameters, which are needed to create a configuration:
98+
99+
100+
<table>
101+
<tr>
102+
<th valign="top">
103+
104+
Parameter
105+
106+
</th>
107+
<th valign="top">
108+
109+
Description
110+
111+
</th>
112+
</tr>
113+
<tr>
114+
<td valign="top">
115+
116+
count
117+
118+
</td>
119+
<td valign="top">
120+
121+
Number of models available description
122+
123+
</td>
124+
</tr>
125+
<tr>
126+
<td valign="top">
127+
128+
description
129+
130+
</td>
131+
<td valign="top">
132+
133+
Description for model
134+
135+
</td>
136+
</tr>
137+
<tr>
138+
<td valign="top">
139+
140+
executableId
141+
142+
</td>
143+
<td valign="top">
144+
145+
Executable under which the model is available, required at the time of creating Deployment configuration model
146+
147+
</td>
148+
</tr>
149+
<tr>
150+
<td valign="top">
151+
152+
model
153+
154+
</td>
155+
<td valign="top">
156+
157+
Required as modelName at the time of creating Deployment configuration versions
158+
159+
</td>
160+
</tr>
161+
<tr>
162+
<td valign="top">
163+
164+
versions
165+
166+
</td>
167+
<td valign="top">
168+
169+
List of available model versions for given model
170+
171+
</td>
172+
</tr>
173+
<tr>
174+
<td valign="top">
175+
176+
versions.name
177+
178+
</td>
179+
<td valign="top">
180+
181+
Optional value to provide as modelVersion at the time of creating Deployment Configuration
182+
183+
</td>
184+
</tr>
185+
<tr>
186+
<td valign="top">
187+
188+
versions.isLatest
189+
190+
</td>
191+
<td valign="top">
192+
193+
Suggests if the modelVersion is latest. See Model Lifecycle for more information
194+
195+
</td>
196+
</tr>
197+
</table>
198+
24199
<a name="task_m15_5kw_tyb"/>
25200
26201
<!-- task\_m15\_5kw\_tyb -->
@@ -29,8 +204,6 @@ You make a model available for use by creating a deployment. You can do so one t
29204
30205
31206
32-
33-
34207
<a name="task_m15_5kw_tyb__steps_n15_5kw_tyb"/>
35208
36209
## Procedure

‎docs/sap-ai-core/create-a-deployment-for-orchestration-4387aa7.md‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -11,7 +11,7 @@
1111
## Prerequisites
1212

1313
> ### Note:
14-
> It is likely that you already have an orchestration deployment running in your default resource group. These steps are only required for users who do not have an orchestration deployment. For more information, see [Get Your Orchestration Deployment URL](get-your-orchestration-deployment-url-ec7c703.md).
14+
> It is likely that you already have an orchestration deployment running in your default resource group. These steps are only required for users who do not have an orchestration deployment or who would like to use model restriction. For more information, see [Get Your Orchestration Deployment URL](get-your-orchestration-deployment-url-ec7c703.md) or [Model Restriction](model-restriction-4d499ee.md).
1515
1616
- You have an SAP AI Core service instance and service key. For more information, see [SAP AI Core Initial Setup Documentation](https://help.sap.com/docs/AI_CORE/2d6c5984063c40a59eda62f4a9135bee/38c4599432d74c1d94e70f7c955a717d.html?locale=en-US&state=PRODUCTION&version=CLOUD).
1717
- You’re using the `extended` service plan. For more information, see [Service Plans](service-plans-c7244c6.md) and [Update a Service Plan](update-a-service-plan-924f892.md).

‎docs/sap-ai-core/create-a-generic-secret-1831845.md‎

Lines changed: 5 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -2,7 +2,7 @@
22

33
# Create a Generic Secret
44

5-
A generic secret gives SAP AI Core authorization to utilize your resource group without exposing your credentials.
5+
A generic secret authorizes SAP AI Core to use your resource group without exposing your credentials.
66

77
<a name="task_i3h_n13_tcc"/>
88

@@ -39,6 +39,8 @@ Generic secrets differ from system secrets, like those for object stores or Dock
3939

4040
To allow the rotation of tenant-wide secrets for long-running deployments without requiring a restart, the deployment must mount the tenant-wide secret. It must also monitor the mounted secret for changes instead of relying on an in-memory copy. When a tenant-wide secret is updated, the tenant must observe the `resourceGroupSecretReplicationStatus` field in the `Get Secret` endpoint to confirm that the secret has been successfully replicated across the required resource groups. For more information, see [Consume Generic Secrets in Executions or Deployments](https://help.sap.com/docs/AI_CORE/2d6c5984063c40a59eda62f4a9135bee/185a3245692542a78bfeff87220410c6.html).
4141

42+
Each tenant can have a maximum of five tenant-wide secrets. If you reach this limit, you'll receive an error message. To free up space, you can delete some tenant-wide secrets as described at [Delete a Generic Secret](delete-a-generic-secret-d5d5187.md). Alternatively, you can submit a ticket to request an increase in your quota.
43+
4244
> ### Tip:
4345
> Generic secrets created at the tenant level automatically propagate to all resource groups. However, if a generic secret with the same name is created at the resource-group level, it replaces the tenant-level secret at the time of creation. The system periodically overrides resource-group level secrets with the corresponding tenant-level secret, but this process may take some time. If a resource-group user creates a secret with the same name as an existing tenant-wide secret, it temporarily overwrites the tenant-wide secret at the resource-group level. This can cause issues, especially for critical operations such as metering.
4446
>
@@ -110,6 +112,8 @@ Generic secrets differ from system secrets, like those for object stores or Dock
110112

111113
To allow the rotation of tenant-wide secrets for long-running deployments without requiring a restart, the deployment must mount the tenant-wide secret. It must also monitor the mounted secret for changes instead of relying on an in-memory copy. When a tenant-wide secret is updated, the tenant must observe the `resourceGroupSecretReplicationStatus` field in the `Get Secret` endpoint to confirm that the secret has been successfully replicated across the required resource groups. For more information, see [Consume Generic Secrets in Executions or Deployments](https://help.sap.com/docs/AI_CORE/2d6c5984063c40a59eda62f4a9135bee/185a3245692542a78bfeff87220410c6.html).
112114

115+
Each tenant can have a maximum of five tenant-wide secrets. If you reach this limit, you'll receive an error message. To free up space, you can delete some tenant-wide secrets as described at [Delete a Generic Secret](delete-a-generic-secret-d5d5187.md). Alternatively, you can submit a ticket to request an increase in your quota.
116+
113117
> ### Tip:
114118
> Generic secrets created at the tenant level automatically propagate to all resource groups. However, if a generic secret with the same name is created at the resource-group level, it replaces the tenant-level secret at the time of creation. The system periodically overrides resource-group level secrets with the corresponding tenant-level secret, but this process may take some time. If a resource-group user creates a secret with the same name as an existing tenant-wide secret, it temporarily overwrites the tenant-wide secret at the resource-group level. This can cause issues, especially for critical operations such as metering.
115119
>

‎docs/sap-ai-core/create-a-prompt-template-declarative-815def5.md‎

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -48,7 +48,7 @@ The following formats are supported:
4848
> content: "Some more {{ ?user_input }}"
4949
> defaults:
5050
> instruction: "default instruction"
51-
> additionalFields:
51+
> additional_fields:
5252
> isDev: true
5353
> validations:
5454
> required: true
@@ -59,9 +59,9 @@ The following formats are supported:
5959
> versions: "*"
6060
> ```
6161
62-
- The `defaults` and `additionalFields` fields are optional.
62+
- The `defaults` and `additional_fields` fields are optional.
6363
64-
- The `additionalFields` field is unstructured and can be used to store metadata or configuration objects.
64+
- The `additional_fields` field is unstructured and can be used to store metadata or configuration objects.
6565
6666
6767
Your template syncs automatically. After a few minutes, you'll be able to verify your template by sending a GET request to the endpoint `{{apiurl}}/lm/promptTemplates`.

‎docs/sap-ai-core/create-a-prompt-template-imperative-92453a7.md‎

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -40,7 +40,7 @@ Create a prompt template by sending a POST request to endpoint <code><code>{{api
4040
> "defaults": {
4141
> "categories": "Finance, Tech, Sports"
4242
> },
43-
> "additionalFields": {
43+
> "additional_fields": {
4444
> "modelParams": {
4545
> "temperature": 0.7,
4646
> "max_tokens": 100
@@ -51,9 +51,9 @@ Create a prompt template by sending a POST request to endpoint <code><code>{{api
5151
> }'
5252
> ```
5353
54-
- The `defaults` and `additionalFields` fields are optional.
54+
- The `defaults` and `additional_fields` fields are optional.
5555
56-
- The `additionalFields` field is unstructured and can be used to store metadata or configuration objects.
56+
- The `additional_fields` field is unstructured and can be used to store metadata or configuration objects.
5757
5858
5959
> ### Output Code:

‎docs/sap-ai-core/deploy-models-dd16e8e.md‎

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -145,15 +145,15 @@
145145
146146
147147
148-
<a name="concept_z3s_f3h_vcc__d96e1165"/>
148+
<a name="concept_z3s_f3h_vcc__d93e1165"/>
149149
150150
## Optional Parameters
151151
152152
The duration of a deployment can be limited using the `ttl` parameter. It takes an integer for quantity, and a single letter to specify units of time. Only minutes \(`m`\), hours \(`h`\) and days \(`d`\), are supported, and values must be natural numbers. For example, `"ttl": "5h"` gives the deployment a duration of 5 hours. `4.5h` and `4h30m` are not valid inputs. If no value is passed, the duration of the deployment if indefinite. Once the duration expires, the deployment is stopped and deleted.
153153
154154
155155
156-
<a name="concept_z3s_f3h_vcc__d96e1196"/>
156+
<a name="concept_z3s_f3h_vcc__d93e1196"/>
157157
158158
## Deployment Quotas
159159

0 commit comments

Comments
 (0)