feat: update outdated libraries

Author: Praveen Kumar Jeyarajan

README.md

@@ -1,6 +1,6 @@
-## GenAI Chat Assistant on AWS
+# GenAI Chat Assistant on AWS
 
-### Introduction
+## Introduction
 
 This demo Chat Assistant application centers around the development of an advanced Chat Assistant using Amazon Bedrock and AWS's serverless GenAI solution. The solution demonstrates a Chat Assistant that utilizes the knowledge of the [Amazon SageMaker Developer Guide](https://docs.aws.amazon.com/sagemaker/latest/dg/gs.html?icmpid=docs_sagemaker_lp/index.html) and [SageMaker instance pricing](https://aws.amazon.com/sagemaker/pricing/). This Chat Assistant serves as an example of the power of Amazon Bedrock in processing and utilizing complex data sets, and it’s capability of converting natural language into Amazon Athena queries. It employs open source tools like LangChain and LLamaIndex to enhance its data processing and retrieval capabilities. The article also highlights the integration of various AWS resources, including Amazon S3 for storage, Amazon Kendra as vector store to support the retrieval augmented generation (RAG), AWS Glue for data preparation, Amazon Athena for efficient querying, Amazon Lambda for serverless computing, and Amazon ECS for container management. These resources collectively enable the Chat Assistant to effectively retrieve and manage content from documents and databases, illustrating the potential of Amazon Bedrock in sophisticated Chat Assistant applications.
 

code/lambda-container/Dockerfile

@@ -1,4 +1,4 @@
-FROM public.ecr.aws/lambda/python:3.11@sha256:ac1ea95dc3a292ee5583e2771bb9e8d21e449aedc5b1fcf93accfae55f049df6
+FROM public.ecr.aws/lambda/python:3.11@sha256:4c04b142654ce3d1b8d705880bf5c720599b5984fd4c751763f09f1682308490
 RUN yum update --security -y && yum clean all
 COPY requirements.txt ${LAMBDA_TASK_ROOT}/
 RUN pip install -r requirements.txt --no-cache-dir

code/lambda-container/requirements.txt

@@ -1,8 +1,8 @@
-boto3>=1.34.61
-llama-index==0.12.28
-llama-index-embeddings-bedrock==0.2.1
-llama-index-llms-bedrock==0.1.9
-langchain==0.3.0
-langchain-community==0.2.19
-sqlalchemy==2.0.28
-PyAthena[SQLAlchemy]
+boto3==1.36.1
+llama-index==0.12.40
+llama-index-embeddings-bedrock==0.5.0
+llama-index-llms-bedrock==0.3.8
+langchain==0.3.25
+langchain-community==0.3.24
+sqlalchemy
+PyAthena[SQLAlchemy]==3.14.1

code/lambda-container/sagemaker_dg_rag.py

@@ -39,6 +39,7 @@ def doc_retrieval(query, llm_model="ClaudeInstant", K=5):
         kclient=Connections.kendra_client,
         top_k=K,
         index_id=Connections.kendra_rawdata_index_id,
+        min_score_confidence=0.1,  # Added parameter to fix validation error
     )
     docs = retriever._get_relevant_documents(query, run_manager=None)
 

code/lambda-container/utils.py

@@ -9,6 +9,9 @@ class InMemoryHistory(BaseChatMessageHistory, BaseModel):
     """In memory implementation of chat message history."""
 
     messages: List[BaseMessage] = Field(default_factory=list)
+    
+    class Config:
+        arbitrary_types_allowed = True
 
     def add_messages(self, messages: List[BaseMessage]) -> None:
         """Add a list of messages to the store"""