top | item 35103125 (no title) RoboTeddy | 3 years ago LLaMA isn't built on RLHF, so it may be necessary to create a more extensive prompt. For example:```You are a super intelligent honest question-answering system.Q: What's 2+2?A: 4Q: What color is the sky?A:``` discuss order hn newest guestbest|3 years ago This is Commodore 64 tier answers lgas|3 years ago That's the point, even C64 tier examples in one shot or few shot learning do wonders for changing the behavior of the model. circuit10|3 years ago That's just an example
guestbest|3 years ago This is Commodore 64 tier answers lgas|3 years ago That's the point, even C64 tier examples in one shot or few shot learning do wonders for changing the behavior of the model. circuit10|3 years ago That's just an example
lgas|3 years ago That's the point, even C64 tier examples in one shot or few shot learning do wonders for changing the behavior of the model.
guestbest|3 years ago
lgas|3 years ago
circuit10|3 years ago