README.md

metadata

dataset_info:
  features:
    - name: text
      dtype: string
  splits:
    - name: train
      num_bytes: 2124117137
      num_examples: 1756174
    - name: test
      num_bytes: 23335122
      num_examples: 14677
    - name: valid
      num_bytes: 19402251
      num_examples: 12734
  download_size: 821240120
  dataset_size: 2166854510
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*
      - split: valid
        path: data/valid-*

Code Contests C++ Dataset

This dataset contains C++ programming solutions extracted from the DeepMind Code Contests dataset.

Source

This dataset is derived from the DeepMind Code Contests dataset, which contains programming problems and solutions from various coding competition platforms.

Original Dataset Citation

@article{li2022competition,
  title={Competition-level code generation with alphacode},
  author={Li, Yujia and Choi, David and Chung, Junyoung and Kushman, Nate and Schrittwieser, Julian and Leblond, R{\'e}mi and Eccles, Tom and Keeling, James and Gimeno, Felix and Lago, Agustin Dal and others},
  journal={Science},
  volume={378},
  number={6624},
  pages={1092--1097},
  year={2022},
  publisher={American Association for the Advancement of Science}
}

Dataset Description

• Format: CSV file with a single 'text' column containing C++ programs
• Number of Programs: 27,520
• Source Languages: Only C++ (language ID 2 from the original dataset)
• Preprocessing:
  • Filtered to include only C++ solutions
  • Extracted program text only
  • Removed duplicate solutions

Usage

import pandas as pd

# Load the dataset
df = pd.read_csv('code_contests_cpp.csv')
print(f"Number of C++ programs: {len(df)}")
print("First program:")
print(df['text'].iloc[0][:200] + "...")  # Print first 200 chars of first program

License

This dataset is derived from the DeepMind Code Contests dataset and is subject to the same license terms as the original dataset. Please refer to the original dataset for licensing information.

Creation

This dataset was created using the script in scripts/process_code_contests.py, which filters the original dataset for C++ solutions and extracts the program text.