MLOps Community: LLMs Mini Summit

Name: MLOps%20Community:%20LLMs%20Mini%20Summit
Uploaded: 2023-11-22

Posted Nov 22, 2023 | Views 1.1K

# LLM Fine-tuning

# Large Language Models

# Weights and Biases

# Virtual Meetup

speakers

Thomas Capelle

ML Engineer @ Weights & Biases

Thomas Capelle is a dedicated Machine Learning Engineer at Weights and Biases, diligently working on the Growth Team. He takes on the responsibility of maintaining the liveliness and timeliness of the wandb/examples repository. His contributions also extend to constructing content on ML-OPS and the application of wandb to the industry, along with developing intriguing aspects of deep learning. Prior to his current role, Thomas used deep learning to resolve short-term forecasting challenges for solar energy at Steady Sun. His interdisciplinary background encompasses Urban Planning, Combinatorial Optimization, Transportation Economics, and Applied Math.

+ Read More

Boris Dayma

CEO @ Craiyon

Boris Dayma is the CEO of Craiyon. He studied in France and Brazil where he received his Masters of Science. Boris has now been working on machine learning for the last 5 years. He is actively involved in the open source ecosystem with the development and training of multiple large models including dalle-mini.

+ Read More

Jonathan Whitaker

AI Researcher @ Data Science Castnet

Jonathan Whitaker is a researcher, educator and consultant currently focused on generative AI. He tends to be happiest tinkering with something technical and then translating that research into a form that is both practical and understandable. You can find some of his work through his blog and YouTube channel "DataScienceCastnet".

+ Read More

Robbie McCorkell

Founding Engineer @ Leap Labs

UK native and founding engineer at Leap Labs. In addition to his MSc in Physics, he has 8 years working in tech leadership across retail, finance, and healthcare. Now he is focused on bringing ML interpretability out of the of academia and into the hands of engineers to improve the dark art of ML model development.

+ Read More

Ben Epstein

Founding Software Engineer @ Galileo

Ben was the machine learning lead for Splice Machine, leading the development of their MLOps platform and Feature Store. He is now a founding software engineer at Galileo (rungalileo.io) focused on building data discovery and data quality tooling for machine learning teams. Ben also works as an adjunct professor at Washington University in St. Louis teaching concepts in cloud computing and big data analytics.

+ Read More

SUMMARY

Deep Dive on LLM Fine-tune In his session, Thomas focuses on understanding the ins and outs of fine-tuning LLMs. We all have a lot of questions during the fine-tuning process. How do you prepare your data? How much data do you need? Do you need to use a high-level API, or can you do this in PyTorch? During this talk, we will try to answer these questions. Thomas will share some tips and tricks on his journey in the LLM fine-tuning landscape. What worked and what did not, and hopefully, you will learn from his experience and the mistakes he made.

A Recipe for Training Large Language Models AI models have become orders of magnitude larger in the last few years. Training such large models presents new challenges, and has been mainly practiced in large companies. In this talk, we tackle best practices for training large models, from early prototype to production.

What The Kaggle 'LLM Science Exam' Competition Can Teach Us About LLMs This competition challenged participants to submit a model capable of answering science-related multiple-choice questions. In doing so it provided a fruitful environment for exploring most of the key techniques and approaches being applied today by anyone building with LLMs. In this talk, we look at some key lessons that this competition can teach us.

Do you really know what your model has learned? Leap Labs demonstrates how data-independent model evaluations represent a paradigm shift in the model development process. All through our dashboard’s beautiful Weights & Biases Weave integration.

+ Read More