NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO · Hugging Face
huggingface.co
NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO · Hugging Face
The CoD prompt instructs highly powered LLMs such as GPT-4 to produce an initial sparse, verbose summary of an article containing only a few entities. It then iteratively identifies 1–3 missing entities and fuses them into a rewrite of the previous summary in the same number of words.
in models with between 2 and 7 billion parameters, new capabilities emerge such as the ability to generate different creative text in formats like poems, code, scripts, musical pieces, emails, and letters, and to answer even open-ended and challenging questions in an informative way.