---
dataset_info:
  features:
  - name: function
    dtype: string
  - name: severity
    dtype: string
  splits:
  - name: train
    num_bytes: 1327573
    num_examples: 2473
  - name: test
    num_bytes: 237962
    num_examples: 437
  download_size: 670552
  dataset_size: 1565535
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
license: mit
task_categories:
- text-classification
language:
- en
tags:
- finance
pretty_name: Severity of Vulnerable Solidity Functions
size_categories:
- 1K<n<10K
---
This dataset combines vulnerable functions (scraped from 5 auditting companies: [Codehawks](https://www.codehawks.com/), [ConsenSys](https://consensys.io/), [Cyfrin](https://www.cyfrin.io/), [Sherlock](https://www.sherlock.xyz/), [Trust Security](https://www.trust-security.xyz/)) and auddited functions with no vulnerabilities (scraped from [Etherscan](https://etherscan.io))

The purpose of the dataset is to enable training of classification models to discriminate between the 4 classes: `none`, `low`, `medium` and `high`.

| Field | Description |
|-|-|
| 1. `function` | Raw solidity code                                                                                 |
| 2. `severity` | Severity of vulnerability ('none', `low`, `medium`, `high`) |

# Data Analysis

<img src="https://huggingface.co/datasets/msc-smart-contract-audition/vulnerability-severity-classification/resolve/main/figures/severity-distribution.png">
<img src="https://huggingface.co/datasets/msc-smart-contract-audition/vulnerability-severity-classification/resolve/main/figures/length-severity-distribution.png">

# Additional Info
- The newline characters are escaped (i.e. `\\n`)