Explore the power of dataframes in MLOps through this comprehensive 57-minute talk by Jay Chia, Co-Founder of Eventual, at MLOps Community Meetup #124. Dive into the critical role of dataframes in building and maintaining datasets for machine learning workflows. Learn how dataframes serve as flexible interfaces for complex data processing, analytics, I/O, and feeding ML training pipelines, using Daft (www.getdaft.io) as a practical example. Gain insights into handling unstructured data, modern ML requirements, and the bare-bones ML stack. Watch a live Daft demo and discover its roadmap, including features like streaming and syntax improvements. Understand how dataframes can simplify MLOps processes and enhance your machine learning projects.
Dataframes Are All You Need: MLOps on Easy Mode
Overview
Syllabus
Join the LLMs in Production Conference Part 2 In Person San Francisco Workshop - Deploy and Scale LLM-based applications today at PM PDT. Location:
[] Introduction to Jay Chia
[] Join the virtual LLMs in Production Conference Part 2 and In Person San Francisco Workshop - Deploy and Scale LLM-based applications today!
[] Dataframes Are All You Need: MLOps on Easy Mode
[] Jay's background
[] Dataframes
[] What is a Dataframe?
[] Different Dataframes
[] Table Stakes
[] Let's talk MLOps
[] "Unstructured" data
[] The bare-bones ML Stack
[] Daft Demo
[] Modern ML requirements
[] The bare-bones ML Stack
[] Daft Roadmap
[] Giving access to the repo notebook
[] Streaming
[] Syntax and library expression
[] S3 bucket and the data frames
[] Trading off latency and throughput in the library
[] Build another data frame Library
[] Daft for Eventual
[] System streaming mode
[] Wrap up
Taught by
MLOps.community
