Buy me a coffee!

docker run --rm -v $(pwd):/package tiivik/lambdazipper MYPACKAGE

Example:

docker run --rm -v $(pwd):/package tiivik/lambdazipper numpy

Result is numpy.zip that can be added as a Lambda Layer or extracted into your existing Lambda deployment package.

Confirmed list of Python 3.8 dependencies that can successfully be built for AWS Lambda with the Dockerfile provided

opencv-python
numpy
Pillow
dlib
sklearn
requests
flask
bpy_lambda

1. Docker environment mimicing AWS Lambda environment. If you are looking for an advanced Lambda environment replication, take a look at lambci/docker-lambda.
2. Packaging script to build AWS Lambda ready packages from Python 3.6 modules.

Premade Docker image can be found on Docker Hub https://cloud.docker.com/repository/docker/tiivik/lambdazipper

FROM amazonlinux:2017.03
RUN yum -y install git \
    python38 \
    python38-pip \
    zip \
    && yum clean all
RUN python3 -m pip install --upgrade pip \
    && python3 -m pip install boto3
RUN yum -y install gcc g++ gcc gcc-c++ cmake
RUN yum -y install python38*

ADD package.sh /
ENTRYPOINT ["/package.sh"]

docker build -t my-lambda .