Releases: aws-solutions/generative-ai-application-builder-on-aws
Releases · aws-solutions/generative-ai-application-builder-on-aws
v2.1.5
v2.1.4
v2.1.3
Fixed
- Fixed a bug where source documents were not displaying when streaming was disabled (#164).
- Fixed a bug where the prompt editing disable feature would result in UI failures (#165).
- Fixed a bug where a new conversation was created whenever a Chat error occurs (#166).
- Fixed annoymized metrics collection by updating to supported timeframe of every 3 hours
Security
- Updated package versions to address security vulnerabilities
v2.1.2
v2.1.1
Changed
- AWS Lambda runtime for Nodejs from 20.x to 22.x
- Updated
pyproject.toml
spec updated to use python 3.12 or higher - AWS SDK (boto3 and Javascript SDK) version updates
- Updated warnings and references documentation when selecting 'BYO Cognito user pool'.
Fixed
- An issue with RAG based deployment, specifically when using Anthropic Claude 3.5 v2 under Amazon Bedrock that caused incorrect prompt to be associated with the configuration.
- An issue related to streaming responses in use case deployments, specifically when using Llama 3.2 under Amazon Bedrock.
Security
- Updated node library versions to address security vulnerabilities
v2.1.0
Added
- Support for using Amazon Bedrock Agents (#49).
- Support for new LLMs available through Amazon Bedrock.
- Support for Bedrock cross-region inference profiles.
- Support for using existing Amazon Cognito user pool configuration when deploying the application and its use cases (#129).
Changed
- Use LCEL to replace LangChain
Chains
in the solution's implementation
Fixed
- Fixed issue when removing a score threshold on an existing use case (#154).
Security
- Updated library versions to address security vulnerabilities
External Contributors
- @majdarbash for #129
v2.0.4
v2.0.3
v2.0.2
v2.0.1
Changed
- With the release of AWS-Solutions-Constructs v2.65.0, the AWS ApiGateway websocket integration with Amazon SQS Queue is available in the library. Hence the implementation has been updated to use this construct.
Fixed
- Issue #131 which caused an issue with rendering non-S3 source URLs from vector store for a RAG based use case.
- Issue #132 where configured Guardrails for Amazon Bedrock to have no effect on a use case's text input validations and output responses.
- Wizard validation failure for SageMaker model selection type, that allowed user to navigate ahead even when the page had failed validations.
- An AWS WAF rule that blocked larger payloads for HTTP POST request to the
/deployments
endpoint. This restricted configuring large system prompts (over 8 KB) for use case cases.