template.yaml

AWSTemplateFormatVersion: '2010-09-09'
Transform: AWS::Serverless-2016-10-31
Description: >
  llmlite-service

  SAM Template for llmlite-service

# More info about Globals: https://github.com/awslabs/serverless-application-model/blob/master/docs/globals.rst
Globals:
  Function:
    Timeout: 600
    MemorySize: 128
    Environment:
      Variables:
        WORKER_CONFIG: !Ref WorkerConfigParameter

Parameters:
  AliasParameter:
    Type: String
    Default: live
  WorkerConfigParameter:
    Type: String
    Description: Sample environment variable
    Default: '{"model": null, "alias": null, "api_base": null, "api_version": "2023-07-01-preview", "debug": false, "temperature": null, "max_tokens": null, "request_timeout": 600, "max_budget": null, "telemetry": true, "drop_params": false, "add_function_to_prompt": false, "headers": null, "save": false, "config": null, "use_queue": false}'

Resources:
  MyUrlFunctionPermissions:
    Type: AWS::Lambda::Permission
    Properties:
      FunctionName: !Ref URL
      Action: lambda:InvokeFunctionUrl
      Principal: "*"
      FunctionUrlAuthType: NONE

  Function:
    Type: AWS::Serverless::Function
    Properties:
      FunctionName: !Sub "${AWS::StackName}-function"
      CodeUri: "./litellm"
      Handler: proxy/lambda.handler
      Runtime: python3.11
      AutoPublishAlias: !Ref AliasParameter
      Architectures:
       - x86_64
      DeploymentPreference:
        Type: AllAtOnce
        Alarms:
          - !Ref NewVersionErrorMetricGreaterThanZeroAlarm

  NewVersionErrorMetricGreaterThanZeroAlarm:
    Type: "AWS::CloudWatch::Alarm"
    Properties:
      AlarmDescription: Lambda Function Error > 0
      ComparisonOperator: GreaterThanThreshold
      Dimensions:
        - Name: Resource
          Value: !Sub "${Function}:live"
        - Name: FunctionName
          Value: !Ref Function
        - Name: ExecutedVersion
          Value: !GetAtt Function.Version.Version
      EvaluationPeriods: 1
      Unit: Count
      MetricName: Errors
      Namespace: AWS/Lambda
      Period: 60
      Statistic: Sum
      Threshold: 0

  URL:
    Type: AWS::Lambda::Url
    DependsOn: FunctionAliaslive
    Properties:
      AuthType: NONE
      Qualifier: live
      TargetFunctionArn: !GetAtt Function.Arn

Outputs:
  FunctionARN:
    Description: "Lambda Function ARN"
    Value: !GetAtt Function.Arn

  FunctionUrl:
    Description: "Lambda Function URL Endpoint"
    Value:
      Fn::GetAtt: URL.FunctionUrl

  FunctionVersion:
    Description: "Lambda Function Version"
    Value: !GetAtt Function.Version.Version
  
  FunctionNewAlarmARN:
    Description: "Lambda Function New Alarm ARN"
    Value: !GetAtt NewVersionErrorMetricGreaterThanZeroAlarm.Arn