five

francepfl/mantis_dataset

收藏
Hugging Face2024-05-21 更新2024-06-12 收录
下载链接:
https://hf-mirror.com/datasets/francepfl/mantis_dataset
下载链接
链接失效反馈
官方服务:
资源简介:
--- dataset_info: - config_name: default features: - name: id dtype: string - name: images list: - name: bytes dtype: 'null' - name: path dtype: string - name: conversation list: - name: content dtype: string - name: role dtype: string splits: - name: train num_bytes: 32321172 num_examples: 23201 - name: eval num_bytes: 8707158 num_examples: 6249 - name: test num_bytes: 19116592 num_examples: 15480 download_size: 2724018 dataset_size: 60144922 - config_name: mantis features: - name: id dtype: string - name: images list: - name: bytes dtype: 'null' - name: path dtype: string - name: conversation list: - name: content dtype: string - name: role dtype: string splits: - name: train num_bytes: 32321172 num_examples: 23201 - name: eval num_bytes: 8707158 num_examples: 6249 - name: test num_bytes: 19116592 num_examples: 15480 download_size: 2724018 dataset_size: 60144922 - config_name: mantis_augmented features: - name: id dtype: string - name: images list: - name: bytes dtype: 'null' - name: path dtype: string - name: conversation list: - name: content dtype: string - name: role dtype: string splits: - name: train num_bytes: 49856315 num_examples: 37245 - name: eval num_bytes: 13410116 num_examples: 10017 - name: test num_bytes: 19116592 num_examples: 15480 download_size: 3309661 dataset_size: 82383023 - config_name: mantis_augmented_cot features: - name: id dtype: string - name: images list: - name: bytes dtype: 'null' - name: path dtype: string - name: conversation list: - name: content dtype: string - name: role dtype: string splits: - name: train num_bytes: 90535982 num_examples: 37245 - name: eval num_bytes: 24348644 num_examples: 10017 - name: test num_bytes: 33638921 num_examples: 15480 download_size: 3661949 dataset_size: 148523547 - config_name: mantis_cot features: - name: id dtype: string - name: images list: - name: bytes dtype: 'null' - name: path dtype: string - name: conversation list: - name: content dtype: string - name: role dtype: string splits: - name: train num_bytes: 54067897 num_examples: 23201 - name: eval num_bytes: 14571634 num_examples: 6249 - name: test num_bytes: 33638921 num_examples: 15480 download_size: 2993150 dataset_size: 102278452 configs: - config_name: default data_files: - split: train path: data/train-* - split: eval path: data/eval-* - split: test path: data/test-* - config_name: mantis data_files: - split: train path: mantis/train-* - split: eval path: mantis/eval-* - split: test path: mantis/test-* - config_name: mantis_augmented data_files: - split: train path: mantis_augmented/train-* - split: eval path: mantis_augmented/eval-* - split: test path: mantis_augmented/test-* - config_name: mantis_augmented_cot data_files: - split: train path: mantis_augmented_cot/train-* - split: eval path: mantis_augmented_cot/eval-* - split: test path: mantis_augmented_cot/test-* - config_name: mantis_cot data_files: - split: train path: mantis_cot/train-* - split: eval path: mantis_cot/eval-* - split: test path: mantis_cot/test-* ---
提供机构:
francepfl
原始信息汇总

数据集概述

配置名称及特征

  1. default

    • 特征:
      • id: 字符串类型
      • images: 列表类型,包含bytes(空类型)和path(字符串类型)
      • conversation: 列表类型,包含content(字符串类型)和role(字符串类型)
    • 分割:
      • train: 32321172字节, 23201个样本
      • eval: 8707158字节, 6249个样本
      • test: 19116592字节, 15480个样本
    • 下载大小: 2724018字节
    • 数据集大小: 60144922字节
  2. mantis

    • 特征与分割与default相同
  3. mantis_augmented

    • 特征与default相同
    • 分割:
      • train: 49856315字节, 37245个样本
      • eval: 13410116字节, 10017个样本
      • test: 19116592字节, 15480个样本
    • 下载大小: 3309661字节
    • 数据集大小: 82383023字节
  4. mantis_augmented_cot

    • 特征与default相同
    • 分割:
      • train: 90535982字节, 37245个样本
      • eval: 24348644字节, 10017个样本
      • test: 33638921字节, 15480个样本
    • 下载大小: 3661949字节
    • 数据集大小: 148523547字节
  5. mantis_cot

    • 特征与default相同
    • 分割:
      • train: 54067897字节, 23201个样本
      • eval: 14571634字节, 6249个样本
      • test: 33638921字节, 15480个样本
    • 下载大小: 2993150字节
    • 数据集大小: 102278452字节

数据文件路径

  • default

    • train: data/train-*
    • eval: data/eval-*
    • test: data/test-*
  • mantis

    • train: mantis/train-*
    • eval: mantis/eval-*
    • test: mantis/test-*
  • mantis_augmented

    • train: mantis_augmented/train-*
    • eval: mantis_augmented/eval-*
    • test: mantis_augmented/test-*
  • mantis_augmented_cot

    • train: mantis_augmented_cot/train-*
    • eval: mantis_augmented_cot/eval-*
    • test: mantis_augmented_cot/test-*
  • mantis_cot

    • train: mantis_cot/train-*
    • eval: mantis_cot/eval-*
    • test: mantis_cot/test-*
5,000+
优质数据集
54 个
任务类型
进入经典数据集
二维码
社区交流群

面向社区/商业的数据集话题

二维码
科研交流群

面向高校/科研机构的开源数据集话题

数据驱动未来

携手共赢发展

商业合作