allenai/olmoe-0125-1b-7b-preference-mix
收藏Hugging Face2025-02-03 更新2025-02-15 收录
下载链接:
https://hf-mirror.com/datasets/allenai/olmoe-0125-1b-7b-preference-mix
下载链接
链接失效反馈官方服务:
资源简介:
OLMoE-1B-7B-0125-Instruct数据集是一个由多个在策略偏好数据集混合而成的数据集,用于DPO训练。它包括了从不同模型生成的366.7k个生成对,这些模型包括Mistral、Tulu、Yi、MPT、Google Gemma、InternLM、Falcon、Qwen、GPT-4、Microsoft Phi和NuMind等。数据集根据ODC-BY许可证发布,适用于研究和教育用途。
The OLMoE-1B-7B-0125-Instruct dataset is a mixture of multiple on-policy preference datasets generated for DPO training. It contains 366.7k generation pairs from various models including Mistral, Tulu, Yi, MPT, Google Gemma, InternLM, Falcon, Qwen, GPT-4, Microsoft Phi, and NuMind. The dataset is released under the ODC-BY license and is intended for research and educational purposes.
提供机构:
allenai



