An open API service indexing awesome lists of open source software.

https://github.com/lixx21/mlops_zoomcamp

All notes and project of MLOPS ZoomCamp 2023
https://github.com/lixx21/mlops_zoomcamp

machine-learning ml-engineer mlops mlops-zoomcamp

Last synced: about 1 year ago
JSON representation

All notes and project of MLOPS ZoomCamp 2023

Awesome Lists containing this project

README

          

# MLOps Zoomcamp from datatalks

All Homework Data: [NYC TRIP DATA](https://www.nyc.gov/site/tlc/about/tlc-trip-record-data.page)

## [01-Intro](/01-Intro/)

    In this module I learned about what is MLOps and the maturity of MLOps. Not just that in this module I also learned about how to setup the environment for our machine (linux).

## [02-Experiment-Tracking](/02-Experiment-Tracking/)

    In this module I learned about how to track our experiment as a data scientist and machine learning engineer. Sometimes as a data scientist we do many of experiments before we get the best model, and we need to tracking those environments, therefore we use ```mlflow``` or ```weight and biases``` to help us tracking our environment an analyzed our experiment (accuracy, RMSE, parameters and so on)

## [03-Orchestation](/03-Orchestration/)

    In this module, I learned about how to use **Prefect** a orchestration tool for MLOps, where we can see our model's peformance every day or schedule our run deployment. To deploy this Prefect we need to push our script to Github. ```data``` ```.prefectignore``` ```deployment.yaml``` ```mlflow.db``` ```prefect.yaml``` and ```mlruns/``` are the dapendencies for prefect deployment using Github, because when we want to deploy Prefect using data, prefect will see the directory from root directory of github. That is why I put all dependencies in main root directory of my github. You can see more of Prefect on [Prefect - Docs](https://docs.prefect.io/2.10.15/)

## [04-Deployment](/04-deployment/)
    In this module, I learned about how to deploy model with several deployment types like, batch and online (streaming and web server)