Continuations of 250 prompts with LLaMA 30B fine-tuned on Open-Assistant data

open-assistant.github.io