Serverless-python-requirements and github actions fail to add python requirements to AWS Lambda

Serverless Framework
,
1

Hi All,

I using Github actions and serverless to deploy a Python AWS Lambda Application. The libraries I have in my requirements.txt are not being added to my lambda functions. I confirmed this by logging the included libraries in my lambda with

installed_packages = pkg_resources.working_set
installed_packages_list = sorted(
   ["%s==%s" % (i.key, i.version) for i in installed_packages]
)

This results in the following log:

image
image1258×376 92.6 KB

Comparing that log with my requirements.txt you can see numpy and scipy are missing:

    boto3==1.16.36
    botocore==1.19.36
    jmespath==0.10.0
    numpy==1.19.4
    python-dateutil==2.8.1
    python-json-logger==2.0.1
    s3transfer==0.3.3
    scipy==1.5.4
    six==1.15.0
    tqdm==4.54.1
    urllib3==1.26.2
    git+ssh://git@github.com/jeb2162/model_scoring.git

My github workflow is below:

name: Model Scoring Service Dev Deployment

on:
  push:
    branches:
      - "**"
      - "!master"


jobs:
  deploy:
    name: deploy
    runs-on: ubuntu-latest
    strategy:
      matrix:
        node-version: [12.x]
    env:
        AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_ID }}
        AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
        DOCKER_USER: ${{ secrets.DOCKER_USER }}
        DOCKER_PASSWORD: ${{ secrets.DOCKER_PASSWORD }}
        STAGE: dev
        SLS_DEBUG: '*'

    steps:
      #- uses: actions/checkout@v1
      - uses: actions/checkout@master

      - name: Use Node.js ${{ matrix.node-version }}
        uses: actions/setup-node@v1
        with:
          node-version: ${{ matrix.node-version }}

      - uses: webfactory/ssh-agent@v0.4.1
        with:
            ssh-private-key: ${{ secrets.SSH_PRIVATE_KEY }}

      - name: Install dependencies
        run: npm install

      - name: Configure AWS Credentials
        uses: aws-actions/configure-aws-credentials@v1
        with:
          aws-access-key-id: ${{ secrets.AWS_ACCESS_KEY_ID }}
          aws-secret-access-key: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
          aws-region: us-east-1
    
      - name: Serverless Deploy 
        run: npm run-script deploy

Here are the relevant parts of my serverless.yml:

service: model-scoring-service

package:
  individually: false
  exclude:
    - v-lambda-env/**
    - node_modules/**
    - package.json
    - package-log.json
  include:
    - requirements.txt

provider:
  name: aws
  runtime: python3.7
  stage: ${env:STAGE}
    
custom:
  datadog:
    flushMetricsToLogs: true
    addLayers: true
    logLevel: "DEBUG"
    enableXrayTracing: false
    enableDDTracing: true
    forwarder: arn:aws:lambda:us-east-1:837515578404:function:datadog-forwarder-Forwarder-1GTYASGLQWO8H
    enableTags: true
    injectLogContext: true

  pythonRequirements:
    dockerizePip: true
    dockerSsh: true
    zip: true
    slim: false
    dockerSshSymlink: ~/.ssh

plugins:
  - serverless-pseudo-parameters
  - serverless-plugin-datadog
  - serverless-python-requirements

Any input/help on why my requirements are getting added to the lambda would be appreciated!