The Chinchilla paper only addresses the contrived use case of a model that is trained once and never used for inference. Since most of the real world compute cost will be in inference, Chinchilla seems to offer little…
I don’t think the switch from “you” to “Copilot” is a hallucination OR a mistake. I think this shows that there are two systems being used together: the original Copilot model, and the chat agent. The chat agent is…
The Chinchilla paper only addresses the contrived use case of a model that is trained once and never used for inference. Since most of the real world compute cost will be in inference, Chinchilla seems to offer little…
I don’t think the switch from “you” to “Copilot” is a hallucination OR a mistake. I think this shows that there are two systems being used together: the original Copilot model, and the chat agent. The chat agent is…