dataset_info: | |
features: | |
- name: input_ids | |
sequence: int32 | |
- name: attention_mask | |
sequence: int8 | |
- name: token_type_ids | |
sequence: int8 | |
- name: labels | |
sequence: int64 | |
- name: pixel_values | |
sequence: | |
sequence: | |
sequence: float32 | |
- name: masked_indices | |
sequence: int64 | |
- name: metadata | |
struct: | |
- name: include_only_masked_tokens_images | |
dtype: bool | |
- name: logic_name | |
dtype: string | |
- name: mask_images_equally | |
dtype: bool | |
splits: | |
- name: train | |
num_bytes: 1095181943 | |
num_examples: 617 | |
download_size: 50697224 | |
dataset_size: 1095181943 | |
configs: | |
- config_name: default | |
data_files: | |
- split: train | |
path: data/train-* | |