README.md

---
license: gpl-3.0
configs:
- config_name: chemtables
  data_files:
  - split: train
    path: chemtables/train-*
  - split: validation
    path: chemtables/validation-*
  - split: test
    path: chemtables/test-*
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
- config_name: discomat
  data_files:
  - split: train
    path: discomat/train-*
  - split: validation
    path: discomat/validation-*
  - split: test
    path: discomat/test-*
- config_name: mltables
  data_files:
  - split: train
    path: mltables/train-*
  - split: validation
    path: mltables/validation-*
  - split: test
    path: mltables/test-*
dataset_info:
- config_name: chemtables
  features:
  - name: paper_id
    dtype: string
  - name: table_id
    dtype: string
  - name: table_code
    dtype: string
  - name: sup_text
    dtype: string
  - name: target_cells
    sequence:
    - name: cell_value
      dtype: string
    - name: cell_raw
      dtype: string
    - name: cell_index
      dtype: string
    - name: cell_row_idx
      dtype: int32
    - name: cell_col_idx
      dtype: int32
  - name: gold_json_records
    sequence:
    - name: cell_index
      dtype: string
    - name: cell_record
      dtype: string
  splits:
  - name: train
    num_bytes: 92180
    num_examples: 9
  - name: validation
    num_bytes: 39374
    num_examples: 3
  - name: test
    num_bytes: 117148
    num_examples: 14
  download_size: 124818
  dataset_size: 248702
- config_name: default
  features:
  - name: paper_id
    dtype: string
  - name: table_id
    dtype: string
  - name: table_code
    dtype: string
  - name: sup_text
    dtype: string
  - name: target_cells
    sequence:
    - name: cell_value
      dtype: string
    - name: cell_raw
      dtype: string
    - name: cell_index
      dtype: string
    - name: cell_row_idx
      dtype: int32
    - name: cell_col_idx
      dtype: int32
  - name: gold_json_records
    sequence:
    - name: cell_index
      dtype: string
    - name: cell_record
      dtype: string
  splits:
  - name: train
    num_bytes: 78484
    num_examples: 9
  - name: validation
    num_bytes: 37457
    num_examples: 3
  - name: test
    num_bytes: 113119
    num_examples: 14
  download_size: 122465
  dataset_size: 229060
- config_name: discomat
  features:
  - name: paper_id
    dtype: string
  - name: table_id
    dtype: string
  - name: table_code
    dtype: string
  - name: sup_text
    dtype: string
  - name: target_cells
    sequence:
    - name: cell_value_processed
      dtype: string
    - name: i
      dtype: int32
    - name: j
      dtype: int32
    - name: k
      dtype: int32
  - name: gold_json_records
    sequence:
    - name: cell_index
      sequence: int32
      length: 3
    - name: cell_record
      dtype: string
  splits:
  - name: train
    num_bytes: 2300237
    num_examples: 500
  - name: validation
    num_bytes: 2300237
    num_examples: 500
  - name: test
    num_bytes: 2366158
    num_examples: 487
  download_size: 1430344
  dataset_size: 6966632
- config_name: mltables
  features:
  - name: paper_id
    dtype: string
  - name: table_id
    dtype: string
  - name: table_code
    dtype: string
  - name: sup_text
    dtype: string
  - name: target_cells
    sequence:
    - name: cell_value
      dtype: string
    - name: cell_raw
      dtype: string
    - name: cell_value_char_idx_start
      dtype: int32
    - name: cell_value_char_idx_end
      dtype: int32
    - name: cell_raw_char_idx_start
      dtype: int32
    - name: cell_raw_char_idx_end
      dtype: int32
  - name: gold_json_records
    sequence:
    - name: cell_char_index
      sequence: int32
      length: 2
    - name: cell_record
      dtype: string
  splits:
  - name: train
    num_bytes: 696651
    num_examples: 43
  - name: validation
    num_bytes: 150816
    num_examples: 11
  - name: test
    num_bytes: 1248693
    num_examples: 68
  download_size: 605737
  dataset_size: 2096160
---