junlinw/opc-sft-s2-annealing-ins3-python-precode0.5-cb-og0.1entire_OG_var3
收藏Hugging Face2025-10-06 更新2025-10-25 收录
下载链接:
https://hf-mirror.com/datasets/junlinw/opc-sft-s2-annealing-ins3-python-precode0.5-cb-og0.1entire_OG_var3
下载链接
链接失效反馈官方服务:
资源简介:
该数据集包含了输入ID(input_ids)、标签(labels)、时间戳(t)、提示符令牌数(num_prompt_tokens)、遮蔽令牌数(num_mask_tokens)和原始数据(original)等字段。数据集分为训练集和测试集,其中训练集包含2,219,937个示例,大小为5,665,027,732字节;测试集包含22,423个示例,大小为60,072,468字节。
The dataset includes fields such as input IDs (input_ids), labels (labels), timestamps (t), number of prompt tokens (num_prompt_tokens), number of masked tokens (num_mask_tokens), and original data (original). The dataset is split into a training set with 2,219,937 examples and a size of 5,665,027,732 bytes, and a test set with 22,423 examples and a size of 60,072,468 bytes.
提供机构:
junlinw



