GenRM/dolphin-r1-cognitivecomputations
收藏官方服务:
资源简介:
Dolphin R1是一个由800k个样本组成的数据集,用于训练R1风格的推理模型。该数据集包括来自DeepSeek-R1的30万个推理样本,来自Gemini 2.0闪存思维的30万个推理样本,以及20万个Dolphin聊天样本。
Dolphin R1 is a dataset composed of 800k samples designed for training R1-style reasoning models. It includes 300k reasoning samples from DeepSeek-R1, 300k reasoning samples from Gemini 2.0 flash thinking, and 200k samples of Dolphin chat.
提供机构:
GenRM


