Новости

Workshop from GlowByte LLC:
"How to Organize a Workspace for Data Science"
MLOps Implementation Expert, Advanced Analytics Practice Employee, Ph.D. Igor Rytsarev

Workshop Description:

The workshop will delve into the topic of organizing a workspace for Data Scientists (and beyond) and will help understand how to streamline work with ML models. 

During the workshop, we will:
- Discuss technologies that simplify the life of a DS.
- Debate on how not to work with projects.
- Learn how to organize model development.
- Achieve experiment reproducibility.
- Familiarize ourselves with model packaging. 

Who Will Benefit from the Workshop?

Model developers (and beyond) who want to learn:
- How to use modern tools.
- How to organize teamwork.
- How to increase model productivity.

Workshop Agenda:

- How to organize a model development environment?
- What are "Virtual Environments" and why are they needed?
- Jupyter (Notebook/Lab/Hub).
- Standard project structure templating.
- Code quality.
- Dataset versioning.
- What is an experiment and how to manage them?
- Using MLFlow to manage experiments.
- Model packaging (Creating a simple API for the model, building images using Docker).

Upon completion of the workshop, an electronic certificate will be issued, which is taken into account when considering resumes in our company.

Prerequisites:

- Experience with Python.
- Experience with git (init/pull/commit/push).
- Personal laptop (with at least 2 processor cores, at least 8 GB RAM, Windows 8 and above (but Unix/MacOS systems are also acceptable)).
- Installed libraries: dvc, mlflow, cookiecutter.
- Installed Anaconda (https://www.anaconda.com/download).
- Installed Git (https://git-scm.com/download).
- Installed Docker (https://www.docker.com/products/docker-desktop).