ZixuanKe/fingpt_convfinqa_sup_sample_from_policy_v1.1_dpo_binarized
收藏Hugging Face2024-11-26 更新2024-12-14 收录
下载链接：
https://hf-mirror.com/datasets/ZixuanKe/fingpt_convfinqa_sup_sample_from_policy_v1.1_dpo_binarized
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含多个字段，主要用于处理与提示（prompt）、被拒绝的响应（rejected）和选择的响应（chosen）相关的数据。每个字段都有对应的长度信息，如llama3_prompt_length、llama3_chosen_length和llama3_rejected_length。数据集分为训练集和验证集，训练集包含28286个样本，验证集包含1362个样本。
This dataset includes multiple fields primarily dealing with data related to prompts, rejected responses, and chosen responses. Each field has corresponding length information such as llama3_prompt_length, llama3_chosen_length, and llama3_rejected_length. The dataset is divided into training and validation sets, with the training set containing 28,286 samples and the validation set containing 1,362 samples.
提供机构：
ZixuanKe
