metadata

language:
  - en
license: mit
size_categories:
  - 1M<n<10M
task_categories:
  - text-generation
paperswithcode_id: disl
pretty_name: 'DISL: Fueling Research with A Large Dataset of Solidity Smart Contracts'
configs:
  - config_name: decomposed
    data_files: data/decomposed/*
  - config_name: raw
    data_files: data/raw/*
tags:
  - code
  - solidity
  - smart contracts
  - webdataset

DISL

The DISL dataset features a collection of 514506 unique Solidity files that have been deployed to Ethereum mainnet. It caters to the need for a large and diverse dataset of real-world smart contracts. DISL serves as a resource for developing machine learning systems and for benchmarking software engineering tools designed for smart contracts.

Content

the raw subset has full contracts source code and it's not deduplicated, it has 3,298,271 smart contracts
the decomposed subset contains Solidity files, it is derived from raw, it is deduplicated using Jaccard similarity with a threshold of 0.9, it has 514,506 Solidity files
the cuttoff date is January 15, 2024, approx at block 19010000

If you use DISL, please cite the following tech report:

@techreport{disl2403.16861,
 title = {DISL: Fueling Research with A Large Dataset of Solidity Smart Contracts},
 year = {2024},
 author = {Gabriele Morello and Mojtaba Eshghie and Sofia Bobadilla and Martin Monperrus},
 url = {http://arxiv.org/pdf/2403.16861},
 number = {2403.16861},
 institution = {arXiv},
}

Curated by: Gabriele Morello

Instructions to explore the dataset

from datasets import load_dataset

# Load the raw dataset
dataset = load_dataset("ASSERT-KTH/DISL", "raw")

# OR

# Load the decomposed dataset
dataset = load_dataset("ASSERT-KTH/DISL", "decomposed")

# number of rows and columns
num_rows = len(dataset["train"])
num_columns = len(dataset["train"].column_names)

# random row
import random
random_row = random.choice(dataset["train"])

# random source code
random_sc = random.choice(dataset["train"])['source_code']
print(random_sc)