search

aws-solutions / generative-ai-application-builder-on-aws

Generative AI Application Builder on AWS facilitates the development, rapid experimentation, and deployment of generative artificial intelligence (AI) applications without requiring deep experience in AI. The solution includes integrations with Amazon Bedrock and its included LLMs, such as Amazon Titan, and pre-built connectors for 3rd-party LLMs.
https://aws.amazon.com/solutions/implementations/generative-ai-application-builder-on-aws
Apache License 2.0
129 stars 38 forks source link

support for query of non-english documents #68

Open michlin0825 opened 4 months ago

michlin0825 commented 4 months ago

Describe the feature you'd like requesting support for inquiry of non-english documents indexed in Kendra, with configuration options during the GenAI application setup.

Additional context I have many non-English documents already indexed in Kendra. I can search those inside Kendra when specifying the default language of the source documents (as in captured screenshot). I wish to achieve the same result using the GenAI application builder, taking advantage of the LLM models in Bedrock.

1_Query_Non_English_Doc_Possible_Inside_Kendra
mukitmomin commented 4 months ago

Hi @michlin0825 that for opening this issue and for using Generative AI Application Builder on AWS.

We are actively looking into the best way to expose knowledge base specific configuration parameters, such as language to the wizard ui. With the release of v1.3.0 with PR https://github.com/aws-solutions/generative-ai-application-builder-on-aws/pull/53 we have added backend support to specify the language for the kendra index. After deploying a chat use case, you can update the AWS Systems Manager (SSM) Parameter Store parameter for the deployment to specify an AttributeFilter JSON object that will be picked up by Kendra.

To locate the SSM parameter, note down the use case ID for your deployment, which is the last 8 digits of the name of the CloudFormation stack for the chat deployment. Navigate to the AWS Console > Systems Manger > Parameter Store, and locate the /gaab-ai/use-case-config/<use-case-id>. Then follow the instructions on our Implementation Guide to specify the language parameter.