AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
Custom Foundation Models for Custom LLMs
In six to 12 months, there'll be competitive models in the low billion parameters that companies can use for domain specific tests. And these are going to be if you look at what's happening on the cloud provider side with the release of Inferentia version two from AWS. With techniques like LORA, which allow an average organization with $500 to $1000 worth of training cost to now fine tune a llama size model in a day or two on a 100s to their specific use case.