Datasets:

PNYX
/

babisteps

language:
  - en
license: mit
task_categories:
  - question-answering
pretty_name: bAbI-steps divided by task/skills
description: TODO
configs:
  - config_name: simpletracking
    data_files:
      - split: test
        path: dataset/simpletracking/simpletracking-test.parquet
      - split: train
        path: dataset/simpletracking/simpletracking-train.parquet
      - split: validation
        path: dataset/simpletracking/simpletracking-validation.parquet
  - config_name: pathfinding
    data_files:
      - split: test
        path: dataset/pathfinding/pathfinding-test.parquet
      - split: train
        path: dataset/pathfinding/pathfinding-train.parquet
      - split: validation
        path: dataset/pathfinding/pathfinding-validation.parquet
  - config_name: listing
    data_files:
      - split: test
        path: dataset/listing/listing-test.parquet
      - split: train
        path: dataset/listing/listing-train.parquet
      - split: validation
        path: dataset/listing/listing-validation.parquet
  - config_name: temporalorder
    data_files:
      - split: test
        path: dataset/temporalorder/temporalorder-test.parquet
      - split: train
        path: dataset/temporalorder/temporalorder-train.parquet
      - split: validation
        path: dataset/temporalorder/temporalorder-validation.parquet
  - config_name: spatialorder
    data_files:
      - split: test
        path: dataset/spatialorder/spatialorder-test.parquet
      - split: train
        path: dataset/spatialorder/spatialorder-train.parquet
      - split: validation
        path: dataset/spatialorder/spatialorder-validation.parquet
  - config_name: sizeorder
    data_files:
      - split: test
        path: dataset/sizeorder/sizeorder-test.parquet
      - split: train
        path: dataset/sizeorder/sizeorder-train.parquet
      - split: validation
        path: dataset/sizeorder/sizeorder-validation.parquet
  - config_name: immediateorder
    data_files:
      - split: test
        path: dataset/immediateorder/immediateorder-test.parquet
      - split: train
        path: dataset/immediateorder/immediateorder-train.parquet
      - split: validation
        path: dataset/immediateorder/immediateorder-validation.parquet
  - config_name: complextracking
    data_files:
      - split: test
        path: dataset/complextracking/complextracking-test.parquet
      - split: train
        path: dataset/complextracking/complextracking-train.parquet
      - split: validation
        path: dataset/complextracking/complextracking-validation.parquet
  - config_name: temporaltracking
    data_files:
      - split: test
        path: dataset/temporaltracking/temporaltracking-test.parquet
      - split: train
        path: dataset/temporaltracking/temporaltracking-train.parquet
      - split: validation
        path: dataset/temporaltracking/temporaltracking-validation.parquet

bAbI-steps Dataset : Per-Task

from datasets import load_dataset
ds = load_dataset('PNYX/babisteps', '<task>')

The available tasks are:

Task ID	Task Name	Split Name
1	simpletracking	train, validation, test
2	immediateorder	train, validation, test
3	complextracking	train, validation, test
4	listing	train, validation, test
5	sizeorder	train, validation, test
6	spatialorder	train, validation, test
7	temporalorder	train, validation, test
8	pathfinding	train, validation, test
9	temporaltracking	train, validation, test

PAPER NAME

@article{TODO, title={TODO}, author={TODO}, journal={TODO}, year={TODO} }