allenai/llama-3.1-tulu-3-70b-preference-mixture
收藏Hugging Face2024-12-02 更新2024-12-14 收录
下载链接：
https://hf-mirror.com/datasets/allenai/llama-3.1-tulu-3-70b-preference-mixture
下载链接
链接失效反馈官方服务：
资源简介：
Llama 3.1 Tulu 3 70B Preference Mixture数据集是一个用于DPO（Direct Preference Optimization）训练的数据集，包含334,367个生成对。这些生成对来自多个不同的模型，如Mistral、Tulu、Yi、MPT、Google Gemma、InternLM、Falcon、Qwen、Llama、GPT-4和Claude等。数据集的主要用途是用于研究目的，并且部分数据受到不同的许可证限制。
This is a preference mixture dataset used for DPO training, specifically for training the Llama 3.1 Tulu 3 70B SFT model to obtain the Llama 3.1 Tulu 3 70B DPO model. The dataset is a mixture of multiple preference datasets, containing 334,367 generation pairs, which were generated using various models. The dataset is licensed under ODC-BY, intended for research and educational use, but some subsets of the dataset may have different licenses, and some portions are non-commercial.
提供机构：
allenai
