Super Data Science: ML & AI Podcast with Jon Krohn

650: SparseGPT: Remove 100 Billion Parameters but Retain 100% Accuracy

Feb 3, 2023
Guest
Jon Krohn
Jon Krohn discusses SparseGPT, a pruning technique for large language models like GPT-3. It can reduce model size by up to 90% without sacrificing accuracy, leading to cost savings. The episode also mentions an upcoming NLP conference and a free access deal for O'Reilly platform.
Ask episode
Chapters
Transcript
Episode notes
1
Exploring the Pruning Technique Sparse GPT for Large Language Models
00:00 • 8min