Large models on CPUs (Practical AI #221)
Model sizes are crazy these days with billions and billions of parameters. As Mark Kurtz explains in this episode, this makes inference slow and expensive despite the fact that up to 90%+ of the parameters don't influence the outputs at all. Mark helps us understand all of the practicalities and progress that is being made in model optimization and CPU inference, including the increasing opportunities to run LLMs and other Generative AI models on commodity hardware.
Embed this episode
Chapters
- 0:44 Neural Magic Mark Kurtz
- 3:24 Why does LLM size matter?
- 6:15 GPUs vs. CPUs
- 8:45 Overcoming perception
- 10:54 Most parameters dont affect results
- 16:01 Balancing space & sparsity
- 17:47 Tackling performance hits
- 20:38 Aware optimization vs not?
- 23:52 Community tools
- 26:11 Neural Magic tools
- 29:56 Supporting new architecture
- 31:40 Exciting research trends
- 34:52 Looking forward in this space
- 37:05 Outro
No transcript yet.
Frequently asked questions
How long is this episode of Changelog Master Feed?
This episode is 38 minutes long.
When was this Changelog Master Feed episode published?
This episode was published on May 2, 2023.
Is there a transcript available for this episode?
Yes, a full transcript is available for this episode. You can read the complete transcript on the episode page.
Can I download this Changelog Master Feed episode?
Yes. Use the download control on the episode player to save the publisher-provided media file.