set the channel description: Make training large models on flyte easy and highly efficient

Folks welcome, So we are in the process of completely overhauling the persistence layer in flytekit. And we are also making many things streamable, would love help in design, testing and even implementation

has renamed the channel from "sig-large-models" to "large-language-models"

@loud-salesmen-31014 has left the channel

This is a very quiet channel. Is that because it all works seamlessly, or no-one is really doing this? I'm about to embark on trying out Flyte for a somewhat large AI company (let's say medium-high volume), and we do image inference, but also video and LLM, and our own internal APIs have struggled to work with both streaming and let's call it "batch" inference.

any code or tutorial to refer how to deploy llama-2 chat using flyte framework