Union at Toronto Machine Learning Summit
Union.ai is excited to share that we will be attending and sponsoring the Toronto Machine Learning Summit conference happening from June 12-13 at The Carlu.
Now in its seventh year, the Toronto Machine Learning Summit (TMLS) conference brings together leading experts, researchers and practitioners from the field of machine learning. Through cutting-edge research, hands-on workshops and industry case-studies, TMLS focuses on community, learning and accessibility.
Union’s participation in this event reflects our commitment to the ML community and ecosystem as we aim to foster knowledge sharing, innovation and collaboration. We plan an exciting virtual workshop on fine-tuning language models with declarative workflow orchestration using Flyte, a workflow orchestrator for data and machine learning. We hope you have a chance to check it out.
Virtual Workshop: Fine-Tuning Language Models and Infrastructure Challenges
Union Chief Machine Learning Engineer Niels Bantilan will conduct a virtual workshop on June 12, from 12 p.m. to 1:30 p.m. EDT. The workshop aims to guide attendees through the process of fine-tuning language models using declarative ML orchestration.
Over the past few years, Large Language Models (LLMs) have become increasingly popular thanks to the availability of datasets and ML frameworks required for training. LLMs serve as foundation models, trained by organizations with significant computational resources. However, adapting these models to specific use cases with proprietary data remains a challenge, mainly because of infrastructure requirements.
During this workshop, we will address:
- The difficulties of setting up the necessary runtime environment for fine-tuning LMs, even when employing cloud tools like Google Colab and consumer-grade GPUs.
- The fundamental concepts behind Flyte, such as containerization for reproducibility, declarative infrastructure, and type-safety.
- How to leverage the latest deep-learning frameworks that optimize memory and computational resources, allowing you to fine-tune language models in a cost-effective and efficient manner.
There’s a lot in store, so we hope you can join us for this workshop! Register here.
Meanwhile, please check out our blog post that dives deeper into the concepts of fine-tuning and prompt engineering and spells out the differences between each approach.
Meet the Team + Giveaway
If you have any questions about the workshop or Flyte, come stop by Booth 9 for answers from our team of ML engineers. We’ll also be hosting a raffle for a one-year subscription to O'Reilly Online Learning so you can expand your skill set and stay up-to-date with the latest industry trends. Feel free to reach out to us on Slack if you plan on attending and would like to set up a 1:1 with one of our team members. We look forward to hearing all about your current ML projects and initiatives.