Custom Foundation Models
The Data Exchange with Ben Lorica
00:00
The Complexity of Deep Learning
Haggai: Big tech companies already use this technology today. He says they train one large model, invests a lot of resources and training is more than refreshing it regularly on your data as it comes in. Haggai: It's surprising how much, how bad is hardware GPU utilization on large language model training. And that's the same concern whether you're a healthcare company, financial company or any other company.
Transcript
Play full episode
