chore(knowledgebase): add support for vector type (#949)

* chore(knowledgebase): add support for vector type

---------

Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>

Author: krokoko

.projen/deps.json

@@ -29,7 +29,7 @@ import {
 const GITHUB_USER = 'awslabs';
 const PUBLICATION_NAMESPACE = 'cdklabs';
 const PROJECT_NAME = 'generative-ai-cdk-constructs';
-const CDK_VERSION: string = '2.177.0';
+const CDK_VERSION: string = '2.178.0';
 
 function camelCaseIt(input: string): string {
   // Hypens and dashes to spaces and then CamelCase...

.projenrc.ts

@@ -1,3 +1,7 @@
+# CDK Generative AI Constructs V0.1.293 (2025-02-10)
+
+Based on CDK library version 2.178.0
+
 # CDK Generative AI Constructs V0.1.291 (2025-01-26)
 
 Based on CDK library version 2.177.0

CHANGELOG.md

@@ -16,7 +16,7 @@ Default output format [None]: json
 ```
 
 - [Node](https://nodejs.org/en) >= v20.9.0
-- [AWS CDK](https://github.com/aws/aws-cdk/releases/tag/v2.177.0) >= 2.177.0
+- [AWS CDK](https://github.com/aws/aws-cdk/releases/tag/v2.178.0) >= 2.178.0
 - [Python](https://www.python.org/downloads/) >=3.9
 - [Projen](https://github.com/projen/projen) >= 0.91.5
 - [Yarn](https://classic.yarnpkg.com/lang/en/docs/cli/install/) >= 1.22.19

DEVELOPER_GUIDE.md

@@ -35,6 +35,7 @@
 - [SharePointDataSourceAuthType](enumerations/SharePointDataSourceAuthType.md)
 - [SharePointObjectType](enumerations/SharePointObjectType.md)
 - [TransformationStep](enumerations/TransformationStep.md)
+- [VectorType](enumerations/VectorType.md)
 
 ## Classes
 

apidocs/namespaces/bedrock/README.md

@@ -65,6 +65,12 @@ The ARN of the Bedrock invokable abstraction.
 
 ***
 
+### supportedVectorType?
+
+> `readonly` `optional` **supportedVectorType**: [`VectorType`](../enumerations/VectorType.md)[]
+
+***
+
 ### supportsAgents
 
 > `readonly` **supportsAgents**: `boolean`

apidocs/namespaces/bedrock/classes/BedrockFoundationModel.md

@@ -0,0 +1,28 @@
+[**@cdklabs/generative-ai-cdk-constructs**](../../../README.md)
+
+***
+
+[@cdklabs/generative-ai-cdk-constructs](../../../README.md) / [bedrock](../README.md) / VectorType
+
+# Enumeration: VectorType
+
+The data type for the vectors when using a model to convert text into vector embeddings.
+The model must support the specified data type for vector embeddings. Floating-point (float32)
+is the default data type, and is supported by most models for vector embeddings. See Supported
+embeddings models for information on the available models and their vector data types.
+
+## Enumeration Members
+
+### BINARY
+
+> **BINARY**: `"BINARY"`
+
+`BINARY` convert the data to binary vector embeddings (less precise, but less costly).
+
+***
+
+### FLOATING\_POINT
+
+> **FLOATING\_POINT**: `"FLOAT32"`
+
+`FLOATING_POINT` convert the data to floating-point (float32) vector embeddings (more precise, but more costly).

apidocs/namespaces/bedrock/enumerations/VectorType.md

@@ -8,6 +8,14 @@
 
 ## Properties
 
+### supportedVectorType?
+
+> `readonly` `optional` **supportedVectorType**: [`VectorType`](../enumerations/VectorType.md)[]
+
+Embeddings models have different supported vector types
+
+***
+
 ### supportsAgents?
 
 > `readonly` `optional` **supportsAgents**: `boolean`

apidocs/namespaces/bedrock/interfaces/BedrockFoundationModelProps.md

@@ -149,3 +149,17 @@ type `VectorCollection`, `RedisEnterpriseVectorStore`,
 ```ts
 - A new OpenSearch Serverless vector collection is created.
 ```
+
+***
+
+### vectorType?
+
+> `readonly` `optional` **vectorType**: [`VectorType`](../enumerations/VectorType.md)
+
+The vector type to store vector embeddings.
+
+#### Default
+
+```ts
+- VectorType.FLOATING_POINT
+```

apidocs/namespaces/bedrock/interfaces/VectorKnowledgeBaseProps.md

@@ -32,6 +32,12 @@ The OpenSearch Vector Collection.
 
 ***
 
+### distanceType
+
+> `readonly` **distanceType**: `string`
+
+***
+
 ### indexName
 
 > `readonly` **indexName**: `string`
@@ -48,6 +54,12 @@ The metadata management fields.
 
 ***
 
+### precision
+
+> `readonly` **precision**: `string`
+
+***
+
 ### vectorDimensions
 
 > `readonly` **vectorDimensions**: `number`