cat-searcher/responses-gemma-1.1-2b-it-split-0-all-hf-rewards-re-sample-reward_mean-importance_weighted-0.25
收藏Hugging Face2024-07-17 更新2024-07-22 收录
下载链接:
https://hf-mirror.com/datasets/cat-searcher/responses-gemma-1.1-2b-it-split-0-all-hf-rewards-re-sample-reward_mean-importance_weighted-0.25
下载链接
链接失效反馈官方服务:
资源简介:
该数据集包含多个特征字段,如prompt(提示)、rewards(奖励)、critiques(批评)等,以及多个生成结果(generate_0到generate_4)。此外,还包含奖励的统计信息,如reward_mean(奖励均值)、reward_var(奖励方差)和reward_gap(奖励差距)。数据集包含一个训练集,共有1578个样本,总大小为14201985字节。
This dataset includes multiple feature fields such as prompt, rewards, critiques, and several generated results (generate_0 to generate_4). It also contains statistical information about rewards, such as reward_mean, reward_var, and reward_gap. The dataset contains a training set with 1578 samples and a total size of 14201985 bytes.
提供机构:
cat-searcher



