Lambda function to access Amazon Bedrock This is a sample Lambda function to access Amazon Bedrock.
Pre-requisites ofr using this is that you need to have Amazon Bedrock available in your AWS Account.
Steps to make it work:
- Create a Basic Lambda function Python 3.11: Author from scratch, Give unction name, Runtime-Python 3.11, Architecture-x86_64, permissions as default to start with. Hit create function
- Add Bedrock Permissions to Lambda: Once Lambda function is created, go to Configuration-->Permission--RoleName(click on role link)-->Permissions-->Add Permissions-->create Inline policy-->Search for Bedrock-->Select Bedrock-->Add All Actions-->All resources-->give policy name-->create policy
- Set Time-out for Lambda to say 1 min: Sometimes it takes time to execute. You can do this in Lambda-->Configuration-->General Configuration
- Upload Lambda Layers zip file to S3: Download zip file from github repo-->almbda layers folder and upload it to S3 in your bucket (same region as Bedrock)
- Create a layer for Bedrock: Go to Lambda Screen. Go to left menu-->Additionla resources-->Layers-->create Layer-->upload from S3-->Give path of step 4, compatible runtime as Python 3.11-->Create
- Add Custom Layer to Lambda: Add layer created in step 5 into Lambda function, using front-end. Open Lambda function-->Click layers-->Add layer-->custom layer-->Select layer from step 5
- Copy code from lambda_code/lambda_function.py to your lambda function code
It should all work now!