Basic machine learning (ML) project to illustrate steps to deploying to AWS Lambda using Docker.
This repo is part of the YouTube video How to Deploy a Python Machine Learning App using Docker + AWS Lambda
Some super basic data science/machine learning project. I basically went on Kaggle, looked up "nlp dataset" and downloaded the first one with sufficient data. Then I copied and pasted old code on here.
Make sure you have conda installed. Then, install the requirements with bash install_requirements.sh.
Finally, you can build the model by running python build_model.py on the terminal.
RUNNING_LOCAL=True python -m app.main
or
RUNNING_LOCAL=True FILENAME=model-dev/data/emotion-labels-test.csv python -m app.main
where FILENAME is any file with text in the header.
deploy-python-ml/
├── app
│ ├── model
│ ├── preprocessing
└── model-dev
└── data
General steps to deploy your code include
- Build your machine learning model and pipeline
- Create/setup a AWS account
- Package your code in a Docker container
- Upload your Docker image to AWS Elastic Container Registry (ECR)
- Create your AWS Lambda to run the ECR image
- Run/test/configure your AWS Lambda
- Deliver your results to others who may need the results
