https://github.com/UCSC-REAL/SimiFeat

https://github.com/UCSC-REAL/SimiFeat

Last synced: 7 days ago
JSON representation

• Host: GitHub
• URL: https://github.com/UCSC-REAL/SimiFeat
• Owner: UCSC-REAL
• License: other
• Created: 2021-10-09T03:38:27.000Z (about 3 years ago)
• Default Branch: main
• Last Pushed: 2023-05-17T07:50:31.000Z (over 1 year ago)
• Last Synced: 2024-08-02T15:29:51.620Z (3 months ago)
• Language: Python
• Size: 2.83 MB
• Stars: 79
• Watchers: 3
• Forks: 15
• Open Issues: 2
• Metadata Files:
  • Readme: README.md
  • License: LICENSE.md

Awesome Lists containing this project

README

        
# Detecting Corrupted Labels Without Training a Model to Predict

**[Update 5/17/2023]** SimiFeat is a module of [Docta](https://github.com/Docta-ai/docta) now!

- A doctor for your data

- An advanced data-centric AI platform that offers a comprehensive range of services aimed at detecting and rectifying issues in your data.

This code is a PyTorch implementation of the [paper](https://arxiv.org/abs/2110.06283): Detecting Corrupted Labels Without Training a Model to Predict 

## Prerequisites

Python 3.6.9

PyTorch 1.7.1

Torchvision 0.8.2

Full list in ./requirements.txt

Datasets will be downloaded to ./data/.

## Run HOC + Vote Based and Rank Based method

On CIFAR-10 .

```

sh ./test_c10_instance.sh  

```

On CIFAR-100

```

sh ./test_c100_instance.sh  

```