weepcat/compute_weights_hh_partial_reward_model_token_by_token
收藏数据链接：
官方服务：
资源简介：
该数据集包含了对话中选择和拒绝的内容及其角色，以及不同模型对于选择和拒绝内容的奖励分数和权重信息。数据集分为训练集，提供了相关的数据文件路径。
The dataset includes the content and roles of chosen and rejected dialogues, as well as reward scores and weight information for chosen and rejected content from different models. The dataset is split into a training set and provides related data file paths.
提供机构：
weepcat
