Re-TACRED: Addressing Shortcomings of the TACRED Dataset

Overview

Re-TACRED

Re-TACRED: Addressing Shortcomings of the TACRED Dataset
George Stoica, Emmanouil Antonios Platanios, and Barnabás Póczos
In Proceedings of the Thirty-fifth AAAI Conference on Artificial Intelligence 2021

Primary Contact: George Stoica. As of Jan 2021, I am no longer at CMU, and the cs.cmu.edu email may no longer work. Please contact me instead at: [email protected].

Changelog

  • 1.0 - Initial dataset release: Data consisted of 105,206 total instances spread across 40 relations.
  • 1.1 - Updated dataset release: After extensive discussion, we have elected to prune Re-TACRED by ~ 14K instances. The new dataset has 91,467 instances, spread across 40 relations. Pruned data consisted of a mixture of messily segmented entities (and corresponding types), or sentences whose relations were ambigious. While this version is smaller, it is cleaner, and better defined.

This repository contains all relevant resources for using Re-TACRED, a new relation extraction dataset.

For details on this work please check out our:

Below we describe the contents of the four repository directories by name.

Re-TACRED

This directory contains version 1.1 of our revised TACRED dataset patches for each split. Due to licensing restrictions, we cannot provide the complete dataset. However, following Alt, Gabryszak, and Hennig (2020), our patch consists of json files mapping TACRED instances by their id to our revised labels.

The original TACRED dataset is available for download from the LDC here. It is free for members, or $25 for non-members.

Applying the patch is simple and only requires replacing each TACRED instance (where applicable) with our revised relation. For convenience, we provide a script for this named apply_patch.py in the Re-TACRED directory. In the script, you only need to replace

tacred_dir = None
save_dir = None

With the path to your TACRED dataset save directory, and the directory where you wish to save the patched data to respectively.

PA-LSTM, C-GCN & SpanBERT

We base our experiments off of the open-source model repositories of:

However, it is not possible to simply pass Re-TACRED to each model repository because each is hardcoded for TACRED. Thus, we must modify certain files to make each model Re-TACRED compatible. To make it as easy as possible, we provide all our altered files in each named model directory (e.g., the provided PA-LSTM directory). All that needs to be done is to replace the corresponding file in our provided directory with the corresponding file in the original model repository. For instance, you may replace SpanBERT's "run_tacred.py" file with our "run_tacred.py" file. Running experiments is equivalent to how it is performed in the original model repositories.

Note that our files also contain certain "quality of life" changes that make running each model more convenient for us. Examples include adding and tracking the test split while training (as opposed to only the dev set).

Owner
George Stoica
PhD ML @ Georgia Tech
George Stoica
A package to predict protein inter-residue geometries from sequence data

trRosetta This package is a part of trRosetta protein structure prediction protocol developed in: Improved protein structure prediction using predicte

Ivan Anishchenko 185 Jan 07, 2023
Official repository for HOTR: End-to-End Human-Object Interaction Detection with Transformers (CVPR'21, Oral Presentation)

Official PyTorch Implementation for HOTR: End-to-End Human-Object Interaction Detection with Transformers (CVPR'2021, Oral Presentation) HOTR: End-to-

Kakao Brain 114 Nov 28, 2022
Release of the ConditionalQA dataset

ConditionalQA Datasets accompanying the paper ConditionalQA: A Complex Reading Comprehension Dataset with Conditional Answers. Disclaimer This dataset

14 Oct 17, 2022
Unofficial PyTorch reimplementation of the paper Swin Transformer V2: Scaling Up Capacity and Resolution

PyTorch reimplementation of the paper Swin Transformer V2: Scaling Up Capacity and Resolution [arXiv 2021].

Christoph Reich 122 Dec 12, 2022
PyElastica is the Python implementation of Elastica, an open-source software for the simulation of assemblies of slender, one-dimensional structures using Cosserat Rod theory.

PyElastica PyElastica is the python implementation of Elastica: an open-source project for simulating assemblies of slender, one-dimensional structure

Gazzola Lab 105 Jan 09, 2023
PyTorch GPU implementation of the ES-RNN model for time series forecasting

Fast ES-RNN: A GPU Implementation of the ES-RNN Algorithm A GPU-enabled version of the hybrid ES-RNN model by Slawek et al that won the M4 time-series

Kaung 305 Jan 03, 2023
This repository holds code and data for our PETS'22 article 'From "Onion Not Found" to Guard Discovery'.

From "Onion Not Found" to Guard Discovery (PETS'22) This repository holds the code and data for our PETS'22 paper titled 'From "Onion Not Found" to Gu

Lennart Oldenburg 3 May 04, 2022
An implementation of an abstract algebra for music tones (pitches).

nbdev template Use this template to more easily create your nbdev project. If you are using an older version of this template, and want to upgrade to

Open Music Kit 0 Oct 10, 2022
AdamW optimizer for bfloat16 models in pytorch.

Image source AdamW optimizer for bfloat16 models in pytorch. Bfloat16 is currently an optimal tradeoff between range and relative error for deep netwo

Alex Rogozhnikov 8 Nov 20, 2022
PyTorch implementation of NeurIPS 2021 paper: "CoFiNet: Reliable Coarse-to-fine Correspondences for Robust Point Cloud Registration"

PyTorch implementation of NeurIPS 2021 paper: "CoFiNet: Reliable Coarse-to-fine Correspondences for Robust Point Cloud Registration"

76 Jan 03, 2023
Turi Create simplifies the development of custom machine learning models.

Quick Links: Installation | Documentation | WWDC 2019 | WWDC 2018 Turi Create Check out our talks at WWDC 2019 and at WWDC 2018! Turi Create simplifie

Apple 10.9k Jan 01, 2023
Package for extracting emotions from social media text. Tailored for financial data.

EmTract: Extracting Emotions from Social Media Text Tailored for Financial Contexts EmTract is a tool that extracts emotions from social media text. I

13 Nov 17, 2022
PyTorch framework for Deep Learning research and development.

Accelerated DL & RL PyTorch framework for Deep Learning research and development. It was developed with a focus on reproducibility, fast experimentati

Catalyst-Team 29 Jul 13, 2022
Image Captioning using CNN and Transformers

Image-Captioning Keras/Tensorflow Image Captioning application using CNN and Transformer as encoder/decoder. In particulary, the architecture consists

24 Dec 28, 2022
Semi-supervised Stance Detection of Tweets Via Distant Network Supervision

SANDS This is an annonymous repository containing code and data necessary to reproduce the results published in "Semi-supervised Stance Detection of T

2 Sep 22, 2022
UpChecker is a simple opensource project to host it fast on your server and check is server up, view statistic, get messages if it is down. UpChecker - just run file and use project easy

UpChecker UpChecker is a simple opensource project to host it fast on your server and check is server up, view statistic, get messages if it is down.

Yan 4 Apr 07, 2022
A model which classifies reviews as positive or negative.

SentiMent Analysis In this project I built a model to classify movie reviews fromn the IMDB dataset of 50K reviews. WordtoVec : Neural networks only w

Rishabh Bali 2 Feb 09, 2022
A Pytorch reproduction of Range Loss, which is proposed in paper 《Range Loss for Deep Face Recognition with Long-Tailed Training Data》

RangeLoss Pytorch This is a Pytorch reproduction of Range Loss, which is proposed in paper 《Range Loss for Deep Face Recognition with Long-Tailed Trai

Youzhi Gu 7 Nov 27, 2021
Official code for Spoken ObjectNet: A Bias-Controlled Spoken Caption Dataset

Official code for our Interspeech 2021 - Spoken ObjectNet: A Bias-Controlled Spoken Caption Dataset [1]*. Visually-grounded spoken language datasets c

Ian Palmer 3 Jan 26, 2022
This is a collection of our NAS and Vision Transformer work.

This is a collection of our NAS and Vision Transformer work.

Microsoft 828 Dec 28, 2022