Dataset Card for Calc-gsm8k

Summary

This dataset is an instance of gsm8k dataset, converted to a simple html-like language that can be easily parsed (e.g. by BeautifulSoup). The data contains 3 types of tags:

• gadget: A tag whose content is intended to be evaluated by calling an external tool (sympy-based calculator in this case)
• output: An output of the external tool
• result: The final answer to the mathematical problem (a number)

Supported Tasks

The dataset is intended for training Chain-of-Thought reasoning models able to use external tools to enhance the factuality of their responses. This dataset presents in-context scenarios where models can outsource the computations in the reasoning chain to a calculator.

Construction Process

The answers in the original dataset were in a structured but non-standard format. So, the answers were parsed, all arithmetical expressions were evaluated using a sympy-based calculator, the outputs were checked to be consistent with the intermediate results and exported into a simple html-like language that BeautifulSoup can parse.

We also perform in-dataset and cross-dataset data-leak detection within the Calc-X collection However, in case of gsm8k, we found no data leaks and removed no examples from the data.

Content and Data splits

For convenience, we created a validation set by sampling 200 random examples from the original train split. This is the default variant:

datasets.load_dataset("MU-NLPC/Calc-gsm8k")

The original data splits can be loaded using:

datasets.load_dataset("MU-NLPC/Calc-gsm8k", "original-splits")

For more info about the content of the dataset, see gsm8k HF dataset and the official repository.

Related work

This dataset was created as a part of a larger effort in training models capable of using a calculator during inference, which we call Calcformers.

• Calc-X collection - datasets for training Calcformers
• Calcformers collection - calculator-using models we trained and published on HF
• Calc-X and Calcformers paper
• Calc-X and Calcformers repo

Here are links to the original dataset:

• original gsm8k dataset
• original gsm8k paper
• original gsm8k repo

Licence

MIT, consistently with the original dataset.

Cite

If you use this version of the dataset in research, please cite the original GSM8K paper, and Calc-X collection as follows:

@inproceedings{kadlcik-etal-2023-soft,
    title = "Calc-X and Calcformers: Empowering Arithmetical Chain-of-Thought through Interaction with Symbolic Systems",
    author = "Marek Kadlčík and Michal Štefánik and Ondřej Sotolář and Vlastimil Martinek",
    booktitle = "Proceedings of the The 2023 Conference on Empirical Methods in Natural Language Processing: Main track",
    month = dec,
    year = "2023",
    address = "Singapore, Singapore",
    publisher = "Association for Computational Linguistics",
    url = "https://arxiv.org/abs/2305.15017",
}