Datasets:

AISE-TUDelft
/

Capybara

configs:
  - config_name: default
    data_files:
      - split: dedup_C
        path: data/dedup_C-*
      - split: dup_C
        path: data/dup_C-*
      - split: dedup_DecomC
        path: data/dedup_DecomC-*
      - split: dup_DecomC
        path: data/dup_DecomC-*
      - split: dedup_demiStripped
        path: data/dedup_demiStripped-*
      - split: dup_demiStripped
        path: data/dup_demiStripped-*
      - split: no_fun_demiStripped
        path: data/no_fun_demiStripped-*
      - split: dup_stripped
        path: data/dup_stripped-*
      - split: dedup_stripped
        path: data/dedup_stripped-*
dataset_info:
  features:
    - name: id
      dtype: int64
    - name: docstring_tokens
      sequence: string
    - name: code_tokens
      sequence: string
    - name: fun_name
      dtype: string
    - name: repo
      dtype: string
    - name: partition
      dtype: string
    - name: opt_level
      dtype: string
  splits:
    - name: dedup_C
      num_bytes: 166445937
      num_examples: 79673
    - name: dup_C
      num_bytes: 345178214
      num_examples: 214587
    - name: dedup_DecomC
      num_bytes: 328727666
      num_examples: 79673
    - name: dup_DecomC
      num_bytes: 610629558
      num_examples: 214587
    - name: dedup_demiStripped
      num_bytes: 315666463
      num_examples: 79673
    - name: dup_demiStripped
      num_bytes: 586705346
      num_examples: 214587
    - name: no_fun_demiStripped
      num_bytes: 603384885
      num_examples: 214587
    - name: dup_stripped
      num_bytes: 60326102
      num_examples: 14245
    - name: dedup_stripped
      num_bytes: 40355724
      num_examples: 7826
  download_size: 585898674
  dataset_size: 3057419895
license: apache-2.0
task_categories:
  - summarization
tags:
  - code
  - Reverse Engineering
  - Binary
  - Code Summarization
size_categories:
  - 100K<n<1M

Dataset Card for "Capybara"

Dataset Description

Repository: https://github.com/AISE-TUDelft/Capybara-BinT5
Paper: https://huggingface.co/papers/2301.01701
Point of Contact: https://huggingface.co/aalkaswan
Raw Data: https://zenodo.org/records/7229809

Dataset Summary

Dataset used to train BinT5. Please refer to the paper for more information.

Citation Information

@inproceedings{alkaswan2023extending,
  title={Extending Source Code Pre-Trained Language Models to Summarise Decompiled Binaries},
  author={Al-Kaswan, Ali and Ahmed, Toufique and Izadi, Maliheh and Sawant, Anand Ashok and Devanbu, Premkumar and van Deursen, Arie},
  booktitle={2023 IEEE International Conference on Software Analysis, Evolution and Reengineering (SANER)},
  pages={260--271},
  year={2023},
  organization={IEEE}
}