math_dataset

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

math_dataset is configured with tfds.core.dataset_builder.BuilderConfig and has the following configurations predefined (defaults to the first one):

  • algebra__linear_1d (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • algebra__linear_1d_composed (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • algebra__linear_2d (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • algebra__linear_2d_composed (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • algebra__polynomial_roots (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • algebra__polynomial_roots_big (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • algebra__polynomial_roots_composed (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • algebra__sequence_next_term (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • algebra__sequence_nth_term (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • arithmetic__add_or_sub (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • arithmetic__add_or_sub_big (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • arithmetic__add_or_sub_in_base (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • arithmetic__add_sub_multiple (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • arithmetic__add_sub_multiple_longer (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • arithmetic__div (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • arithmetic__div_big (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • arithmetic__mixed (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • arithmetic__mixed_longer (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • arithmetic__mul (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • arithmetic__mul_big (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • arithmetic__mul_div_multiple (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • arithmetic__mul_div_multiple_longer (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • arithmetic__nearest_integer_root (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • arithmetic__simplify_surd (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • calculus__differentiate (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • calculus__differentiate_composed (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • comparison__closest (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • comparison__closest_composed (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • comparison__closest_more (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • comparison__kth_biggest (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • comparison__kth_biggest_composed (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • comparison__kth_biggest_more (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • comparison__pair (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • comparison__pair_composed (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • comparison__sort (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • comparison__sort_composed (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • comparison__sort_more (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • measurement__conversion (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • measurement__time (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • numbers__base_conversion (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • numbers__div_remainder (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • numbers__div_remainder_composed (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • numbers__gcd (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • numbers__gcd_composed (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • numbers__is_factor (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • numbers__is_factor_composed (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • numbers__is_prime (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • numbers__is_prime_composed (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • numbers__lcm (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • numbers__lcm_composed (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • numbers__list_prime_factors (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • numbers__list_prime_factors_composed (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • numbers__place_value (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • numbers__place_value_big (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • numbers__place_value_composed (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • numbers__round_number (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • numbers__round_number_big (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • numbers__round_number_composed (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • polynomials__add (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • polynomials__coefficient_named (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • polynomials__collect (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • polynomials__compose (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • polynomials__evaluate (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • polynomials__evaluate_composed (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • polynomials__expand (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • polynomials__simplify_power (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • probability__swr_p_level_set (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • probability__swr_p_level_set_more_samples (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • probability__swr_p_sequence (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

  • probability__swr_p_sequence_more_samples (v1.0.0) (Size: ?? GiB): Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

math_dataset/algebra__linear_1d

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/algebra__linear_1d_composed

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/algebra__linear_2d

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/algebra__linear_2d_composed

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/algebra__polynomial_roots

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/algebra__polynomial_roots_big

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/algebra__polynomial_roots_composed

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/algebra__sequence_next_term

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/algebra__sequence_nth_term

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/arithmetic__add_or_sub

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/arithmetic__add_or_sub_big

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/arithmetic__add_or_sub_in_base

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/arithmetic__add_sub_multiple

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/arithmetic__add_sub_multiple_longer

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/arithmetic__div

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/arithmetic__div_big

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/arithmetic__mixed

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/arithmetic__mixed_longer

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/arithmetic__mul

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/arithmetic__mul_big

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/arithmetic__mul_div_multiple

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/arithmetic__mul_div_multiple_longer

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/arithmetic__nearest_integer_root

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/arithmetic__simplify_surd

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/calculus__differentiate

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/calculus__differentiate_composed

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/comparison__closest

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/comparison__closest_composed

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/comparison__closest_more

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/comparison__kth_biggest

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/comparison__kth_biggest_composed

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/comparison__kth_biggest_more

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/comparison__pair

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/comparison__pair_composed

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/comparison__sort

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/comparison__sort_composed

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/comparison__sort_more

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/measurement__conversion

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/measurement__time

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/numbers__base_conversion

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/numbers__div_remainder

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/numbers__div_remainder_composed

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/numbers__gcd

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/numbers__gcd_composed

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/numbers__is_factor

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/numbers__is_factor_composed

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/numbers__is_prime

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/numbers__is_prime_composed

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/numbers__lcm

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/numbers__lcm_composed

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/numbers__list_prime_factors

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/numbers__list_prime_factors_composed

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/numbers__place_value

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/numbers__place_value_big

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/numbers__place_value_composed

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/numbers__round_number

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/numbers__round_number_big

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/numbers__round_number_composed

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/polynomials__add

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/polynomials__coefficient_named

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/polynomials__collect

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/polynomials__compose

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/polynomials__evaluate

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/polynomials__evaluate_composed

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/polynomials__expand

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/polynomials__simplify_power

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/probability__swr_p_level_set

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/probability__swr_p_level_set_more_samples

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/probability__swr_p_sequence

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

math_dataset/probability__swr_p_sequence_more_samples

Mathematics database.

This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty. This is designed to test the mathematical learning and algebraic reasoning skills of learning models.

Original paper: Analysing Mathematical Reasoning Abilities of Neural Models (Saxton, Grefenstette, Hill, Kohli).

Example usage: train_examples, val_examples = tfds.load( 'mathdataset/arithmetic_mul', split=['train', 'test'], as_supervised=True)

Versions:

  • 1.0.0 (default):

Statistics

None computed

Features

FeaturesDict({
    'answer': Text(shape=(), dtype=tf.string),
    'question': Text(shape=(), dtype=tf.string),
})

Homepage

Supervised keys (for as_supervised=True)

(u'question', u'answer')

Citation

@article{2019arXiv,
  author = {Saxton, Grefenstette, Hill, Kohli},
  title = {Analysing Mathematical Reasoning Abilities of Neural Models},
  year = {2019},
  journal = {arXiv:1904.01557}
}