--- dataset_info: features: - name: example_id dtype: string - name: task dtype: class_label: names: '0': col_type '1': ent_link '2': fetaqa '3': hitab '4': hybridqa '5': infotabs '6': merged_cell_detection '7': rel_extraction '8': row_column_extraction '9': struct_aware_parse '10': tabfact '11': table_cell_extraction '12': table_cell_location '13': table_instruction '14': table_recognition '15': table_size_detection '16': tabmwp '17': tat-qa '18': totto '19': wikibio '20': wikitq - name: src_example_ids dtype: string - name: table_id dtype: string - name: table_seed_id dtype: string - name: table_seed_dataset dtype: string - name: table_page_title dtype: string - name: table_section_title dtype: string - name: table_variant dtype: string - name: img_source dtype: class_label: names: '0': PubTabNet '1': TABMWP '2': seed_render '3': wikipedia - name: input dtype: large_string - name: output dtype: string - name: split dtype: class_label: names: '0': dev '1': test '2': train - name: metadata dtype: string - name: table_wiki_page_id dtype: string - name: table_wiki_old_id dtype: string - name: table_html dtype: large_string - name: table_img dtype: image splits: - name: wikibio_train num_bytes: 11073920562 num_examples: 140000 - name: tatqa_train num_bytes: 114858299 num_examples: 2201 - name: hybridqa_train num_bytes: 14932215002 num_examples: 62670 - name: table_instruction_train num_bytes: 12180312819 num_examples: 136944 - name: tabmwp_train num_bytes: 326184930 num_examples: 23059 - name: hitab_train num_bytes: 1026591527 num_examples: 7417 - name: table_recognition_train num_bytes: 788489258 num_examples: 6927 - name: table_cell_extraction_train num_bytes: 2102295132 num_examples: 7727 - name: table_size_detection_train num_bytes: 1695496703 num_examples: 7800 - name: merged_cell_detection_train num_bytes: 993975274 num_examples: 7500 - name: table_cell_location_train num_bytes: 1227016660 num_examples: 7708 - name: tabfact_train num_bytes: 10356555069 num_examples: 87717 - name: totto_train num_bytes: 36076773336 num_examples: 110934 - name: row_column_extraction_train num_bytes: 2261383645 num_examples: 7721 - name: infotabs_train num_bytes: 1486818637 num_examples: 16538 - name: wikitq_train num_bytes: 3661437235 num_examples: 14152 - name: struct_aware_parse_train num_bytes: 2711066295 num_examples: 126581 - name: fetaqa_train num_bytes: 366072076 num_examples: 3006 download_size: 97565644868 dataset_size: 103381462459 configs: - config_name: default data_files: - split: wikibio_train path: data/wikibio_train-* - split: tatqa_train path: data/tatqa_train-* - split: hybridqa_train path: data/hybridqa_train-* - split: table_instruction_train path: data/table_instruction_train-* - split: tabmwp_train path: data/tabmwp_train-* - split: hitab_train path: data/hitab_train-* - split: fetaqa_train path: data/fetaqa_train-* - split: table_recognition_train path: data/table_recognition_train-* - split: table_cell_extraction_train path: data/table_cell_extraction_train-* - split: table_size_detection_train path: data/table_size_detection_train-* - split: merged_cell_detection_train path: data/merged_cell_detection_train-* - split: table_cell_location_train path: data/table_cell_location_train-* - split: tabfact_train path: data/tabfact_train-* - split: totto_train path: data/totto_train-* - split: row_column_extraction_train path: data/row_column_extraction_train-* - split: infotabs_train path: data/infotabs_train-* - split: wikitq_train path: data/wikitq_train-* - split: struct_aware_parse_train path: data/struct_aware_parse_train-* --- ### TABLET-Small This is the _Small_ sized **train set** of the **TABLET** dataset. It contains the train examples for 14 **TABLET** tasks. Each task is capped at **140,000 examples**, resulting in a total of **776,602 training examples** across **14 tasks**. This dataset is self-contained, each example includes a table image, its HTML representation, and the associated task data. However, if you're interested in downloading just the TABLET tables, check out [TABLET-tables](https://huggingface.co/datasets/alonsoapp/TABLET-tables). All TABLET Subsets: - _(train)_ [**TABLET-Small**](https://huggingface.co/datasets/alonsoapp/TABLET-Small): The smallest TABLET subset, including **776,602 examples** across **14 tasks**. - _(train)_ [**TABLET-Medium**](https://huggingface.co/datasets/alonsoapp/TABLET-Medium): Includes all examples from _TABLET-Small_, plus **Column Type Annotation**, **Entity Linking**, and **Relation Extraction** tasks. Each task is capped at **140,000 examples**, resulting in a total of **1,117,217 training examples** across **17 tasks**. - _(train)_ [**TABLET-Large**](https://huggingface.co/datasets/alonsoapp/TABLET-Large): Includes all examples from _TABLET-Medium_ with **no cap** on task size, resulting in a total of **3,505,311 training examples** across **17 tasks**. - _(dev)_ [**TABLET-dev**](https://huggingface.co/datasets/alonsoapp/TABLET-dev): The **development** set of TABLET. - _(test)_ [**TABLET-test**](https://huggingface.co/datasets/alonsoapp/TABLET-test): The **test** set of TABLET. For more information, see our [paper](https://arxiv.org/pdf/2509.21205), [website](https://precious-panda-5ce815.netlify.app/tablet/), and [GitHub repository](https://github.com/AlonsoApp/TABLET). #### Using the Dataset Given its size, we recommend [streaming](https://huggingface.co/docs/datasets/stream) the dataset instead of downloading it entirely to disk: ```python from datasets import load_dataset dataset = load_dataset('alonsoapp/TABLET-Small', split='fetaqa_train', streaming=True) print(next(iter(dataset))) ``` #### Data Fields Each sample within the dataset is structured with the following fields: * **`example_id`**: Unique identifier for the example. * **`task`**: The name of the task this example belongs to. * **`src_example_ids`**: IDs of the original examples from the source dataset, formatted as `{"Dataset name": "id"}`. Use the `get_original_example` helper function from [our published code](https://github.com/AlonsoApp/TABLET) to easily retrieve the source example. * **`table_id`**: Unique identifier for the table. * **`table_seed_id`**: ID referencing the table in its original (seed) dataset. * **`table_seed_dataset`**: Name of the dataset where the table originated, typically matching the source dataset of the example. * **`table_page_title`**: For tables sourced from Wikipedia, the corresponding page title. * **`table_section_title`**: For Wikipedia tables, the title of the section where the table appears. * **`table_variant`**: Either "raw" or "highlighted". Some examples visually highlight specific cells and this field indicates whether the table is unmodified (raw) or includes highlights (highlighted). * **`img_source`**: Source of the table image. That is, whether the image comes from the Wikipedia visualization (wikipedia), a synthetic renderization from the data in the soruce dataset (seed_render), or directly copied from the original visualization of the table of the source dataset (PubTabNet, TabMWP). * **`input`**: The _instructified_ input used for training and evaluation (see [paper](https://arxiv.org/pdf/2509.21205)). The input can be rephrased using information in `metadata`. * **`output`**: The expected model output for the given `input`. * **`split`**: Dataset split: `train`, `dev`, or `test`. * **`metadata`**: Atomic data for the example to enable reconstruction or rephrasing of the instruction. Each key indicates a data element, the value can be obtained from either the _'input'_ or the _'output'_ strings using the substring defined by the character indexes in 'idx'. Use the get_metadata helper function from [our published code](https://github.com/AlonsoApp/TABLET) to retrieve these values. * **`table_wiki_page_id`**: For Wikipedia tables, the page ID corresponding to the article containing the table (useful for Wikipedia API queries). * **`table_wiki_old_id`**: For Wikipedia tables, the “old ID” identifying the article version at the crawl time. * **`table_html`**: HTML representation of the table. Use the `render_table` helper function from [our code](https://github.com/AlonsoApp/TABLET) to render it in its original style. For highlighted variants, highlighted cells use the CSS class `demeter_highlighted_cell`. Remove any decorators for this class in the CSS to render identically to the raw version. * **`table_img`**: The image representation of the table. #### Citation If you find **TABLET** useful in your research, please consider citing it by the following BibTeX entry. ```bibtex @misc{alonso2025tabletlargescaledatasetrobust, title={TABLET: A Large-Scale Dataset for Robust Visual Table Understanding}, author={Iñigo Alonso and Imanol Miranda and Eneko Agirre and Mirella Lapata}, year={2025}, eprint={2509.21205}, archivePrefix={arXiv}, primaryClass={cs.CV}, url={https://arxiv.org/abs/2509.21205}, } ```