tulu-v3.1-mix-preview-4096-OLMoE
收藏资源简介:
该数据集包含一个名为'messages'的特征,该特征是一个列表,包含'content'和'role'两个子特征,均为字符串类型。数据集分为训练集,包含608042个样本,总大小为1741315705字节。数据集的下载大小为851954240字节。数据集的许可证为odc-by。数据集是通过混合多个其他数据集创建的,包括'allenai/tulu-v2-sft-mixture-olmo-4096'、'HuggingFaceH4/no_robots'、'meta-math/MetaMathQA'、'm-a-p/CodeFeedback-Filtered-Instruction'和'ai2-adapt-dev/daring-anteater-specialized',最大序列长度为4096。数据集还包含一个重命名代码,用于将'OLMo'替换为'OLMoE'。
This dataset includes a feature named `messages`, which is a list comprising two sub-features: `content` and `role`, both of string data type. The dataset is split into a training set containing 608,042 samples with a total size of 1,741,315,705 bytes. The download size of the dataset is 851,954,240 bytes. The dataset is licensed under odc-by. It is constructed by mixing multiple other datasets, including `allenai/tulu-v2-sft-mixture-olmo-4096`, `HuggingFaceH4/no_robots`, `meta-math/MetaMathQA`, `m-a-p/CodeFeedback-Filtered-Instruction`, and `ai2-adapt-dev/daring-anteater-specialized`, with a maximum sequence length of 4096. The dataset also includes a renaming script for replacing `OLMo` with `OLMoE`.
数据集概述
数据集信息
- 特征:
messages: 包含以下子特征的列表content: 数据类型为字符串role: 数据类型为字符串
- 分割:
train: 包含608042个样本,占用1741315705字节
- 下载大小: 851954240字节
- 数据集大小: 1741315705字节
- 配置:
default: 数据文件路径为data/train-*
- 许可证: odc-by
数据集混合配置
- 混合比例:
allenai/tulu-v2-sft-mixture-olmo-4096: 1.0HuggingFaceH4/no_robots: 1.0meta-math/MetaMathQA: 0.25m-a-p/CodeFeedback-Filtered-Instruction: 1.0ai2-adapt-dev/daring-anteater-specialized: 1.0
- 最大序列长度: 4096
相关数据集
| 版本 | 名称 | 摘要 | 最大长度 | 模型名称 |
|---|---|---|---|---|
| v1 | allenai/tulu-v1-sft-mixture | |||
| v2 | allenai/tulu-v2-sft-mixture | 4096 | ||
| v2 | allenai/tulu-v2-sft-mixture-olmo-2048 | 2048 | OLMo-2048 | |
| v3.0 | allenai/tulu-v3.0-mix-preview-4096-OLMo | Tulu 2 + Math/Code + No Robots | 4096 | OLMo |
| v3.0 | allenai/tulu-v3.0-mix-preview-4096-OLMoE | OLMoE Name | 4096 | OLMoE |
| v3.1 | allenai/tulu-v3.1-mix-preview-4096-OLMoE | Add specialized Daring Anteater | 4096 | OLMoE |




