SPRING is a seq2seq model for Text-to-AMR and AMR-to-Text (AAAI2021).

Overview

SPRING

PWC

PWC

PWC

PWC

This is the repo for SPRING (Symmetric ParsIng aNd Generation), a novel approach to semantic parsing and generation, presented at AAAI 2021.

With SPRING you can perform both state-of-the-art Text-to-AMR parsing and AMR-to-Text generation without many cumbersome external components. If you use the code, please reference this work in your paper:

@inproceedings{bevilacqua-etal-2021-one,
    title = {One {SPRING} to Rule Them Both: {S}ymmetric {AMR} Semantic Parsing and Generation without a Complex Pipeline},
    author = {Bevilacqua, Michele and Blloshmi, Rexhina and Navigli, Roberto},
    booktitle = {Proceedings of AAAI},
    year = {2021}
}

Pretrained Checkpoints

Here we release our best SPRING models which are based on the DFS linearization.

Text-to-AMR Parsing

AMR-to-Text Generation

If you need the checkpoints of other experiments in the paper, please send us an email.

Installation

cd spring
pip install -r requirements.txt
pip install -e .

The code only works with transformers < 3.0 because of a disrupting change in positional embeddings. The code works fine with torch 1.5. We recommend the usage of a new conda env.

Train

Modify config.yaml in configs. Instructions in comments within the file. Also see the appendix.

Text-to-AMR

python bin/train.py --config configs/config.yaml --direction amr

Results in runs/

AMR-to-Text

python bin/train.py --config configs/config.yaml --direction text

Results in runs/

Evaluate

Text-to-AMR

python bin/predict_amrs.py \
    --datasets <AMR-ROOT>/data/amrs/split/test/*.txt \
    --gold-path data/tmp/amr2.0/gold.amr.txt \
    --pred-path data/tmp/amr2.0/pred.amr.txt \
    --checkpoint runs/<checkpoint>.pt \
    --beam-size 5 \
    --batch-size 500 \
    --device cuda \
    --penman-linearization --use-pointer-tokens

gold.amr.txt and pred.amr.txt will contain, respectively, the concatenated gold and the predictions.

To reproduce our paper's results, you will also need need to run the BLINK entity linking system on the prediction file (data/tmp/amr2.0/pred.amr.txt in the previous code snippet). To do so, you will need to install BLINK, and download their models:

git clone https://github.com/facebookresearch/BLINK.git
cd BLINK
pip install -r requirements.txt
sh download_blink_models.sh
cd models
wget http://dl.fbaipublicfiles.com/BLINK//faiss_flat_index.pkl
cd ../..

Then, you will be able to launch the blinkify.py script:

python bin/blinkify.py \
    --datasets data/tmp/amr2.0/pred.amr.txt \
    --out data/tmp/amr2.0/pred.amr.blinkified.txt \
    --device cuda \
    --blink-models-dir BLINK/models

To have comparable Smatch scores you will also need to use the scripts available at https://github.com/mdtux89/amr-evaluation, which provide results that are around ~0.3 Smatch points lower than those returned by bin/predict_amrs.py.

AMR-to-Text

python bin/predict_sentences.py \
    --datasets <AMR-ROOT>/data/amrs/split/test/*.txt \
    --gold-path data/tmp/amr2.0/gold.text.txt \
    --pred-path data/tmp/amr2.0/pred.text.txt \
    --checkpoint runs/<checkpoint>.pt \
    --beam-size 5 \
    --batch-size 500 \
    --device cuda \
    --penman-linearization --use-pointer-tokens

gold.text.txt and pred.text.txt will contain, respectively, the concatenated gold and the predictions. For BLEU, chrF++, and Meteor in order to be comparable you will need to tokenize both gold and predictions using JAMR tokenizer. To compute BLEU and chrF++, please use bin/eval_bleu.py. For METEOR, use https://www.cs.cmu.edu/~alavie/METEOR/ . For BLEURT don't use tokenization and run the eval with https://github.com/google-research/bleurt. Also see the appendix.

Linearizations

The previously shown commands assume the use of the DFS-based linearization. To use BFS or PENMAN decomment the relevant lines in configs/config.yaml (for training). As for the evaluation scripts, substitute the --penman-linearization --use-pointer-tokens line with --use-pointer-tokens for BFS or with --penman-linearization for PENMAN.

License

This project is released under the CC-BY-NC-SA 4.0 license (see LICENSE). If you use SPRING, please put a link to this repo.

Acknowledgements

The authors gratefully acknowledge the support of the ERC Consolidator Grant MOUSSE No. 726487 and the ELEXIS project No. 731015 under the European Union’s Horizon 2020 research and innovation programme.

This work was supported in part by the MIUR under the grant "Dipartimenti di eccellenza 2018-2022" of the Department of Computer Science of the Sapienza University of Rome.

Owner
Sapienza NLP group
The NLP group at the Sapienza University of Rome
Sapienza NLP group
An original implementation of "Noisy Channel Language Model Prompting for Few-Shot Text Classification"

Channel LM Prompting (and beyond) This includes an original implementation of Sewon Min, Mike Lewis, Hannaneh Hajishirzi, Luke Zettlemoyer. "Noisy Cha

Sewon Min 92 Jan 07, 2023
RAFT-Stereo: Multilevel Recurrent Field Transforms for Stereo Matching

RAFT-Stereo: Multilevel Recurrent Field Transforms for Stereo Matching This repository contains the source code for our paper: RAFT-Stereo: Multilevel

Princeton Vision & Learning Lab 328 Jan 09, 2023
Gans-in-action - Companion repository to GANs in Action: Deep learning with Generative Adversarial Networks

GANs in Action by Jakub Langr and Vladimir Bok List of available code: Chapter 2: Colab, Notebook Chapter 3: Notebook Chapter 4: Notebook Chapter 6: C

GANs in Action 914 Dec 21, 2022
The code for our paper CrossFormer: A Versatile Vision Transformer Based on Cross-scale Attention.

CrossFormer This repository is the code for our paper CrossFormer: A Versatile Vision Transformer Based on Cross-scale Attention. Introduction Existin

cheerss 238 Jan 06, 2023
Python scripts for performing stereo depth estimation using the HITNET Tensorflow model.

HITNET-Stereo-Depth-estimation Python scripts for performing stereo depth estimation using the HITNET Tensorflow model from Google Research. Stereo de

Ibai Gorordo 76 Jan 02, 2023
Run Keras models in the browser, with GPU support using WebGL

**This project is no longer active. Please check out TensorFlow.js.** The Keras.js demos still work but is no longer updated. Run Keras models in the

Leon Chen 4.9k Dec 29, 2022
The official implementation of ELSA: Enhanced Local Self-Attention for Vision Transformer

ELSA: Enhanced Local Self-Attention for Vision Transformer By Jingkai Zhou, Pich

DamoCV 87 Dec 19, 2022
CRISCE: Automatically Generating Critical Driving Scenarios From Car Accident Sketches

CRISCE: Automatically Generating Critical Driving Scenarios From Car Accident Sketches This document describes how to install and use CRISCE (CRItical

Chair of Software Engineering II, Uni Passau 2 Feb 09, 2022
An implementation of a discriminant function over a normal distribution to help classify datasets.

CS4044D Machine Learning Assignment 1 By Dev Sony, B180297CS The question, report and source code can be found here. Github Repo Solution 1 Based on t

Dev Sony 6 Nov 09, 2021
🥇Samsung AI Challenge 2021 1등 솔루션입니다🥇

MoT - Molecular Transformer Large-scale Pretraining for Molecular Property Prediction Samsung AI Challenge for Scientific Discovery This repository is

Jungwoo Park 44 Dec 03, 2022
Code for ICCV2021 paper SPEC: Seeing People in the Wild with an Estimated Camera

SPEC: Seeing People in the Wild with an Estimated Camera [ICCV 2021] SPEC: Seeing People in the Wild with an Estimated Camera, Muhammed Kocabas, Chun-

Muhammed Kocabas 187 Dec 26, 2022
Keyhole Imaging: Non-Line-of-Sight Imaging and Tracking of Moving Objects Along a Single Optical Path

Keyhole Imaging Code & Dataset Code associated with the paper "Keyhole Imaging: Non-Line-of-Sight Imaging and Tracking of Moving Objects Along a Singl

Stanford Computational Imaging Lab 20 Feb 03, 2022
RM Operation can equivalently convert ResNet to VGG, which is better for pruning; and can help RepVGG perform better when the depth is large.

RM Operation can equivalently convert ResNet to VGG, which is better for pruning; and can help RepVGG perform better when the depth is large.

184 Jan 04, 2023
Autonomous Perception: 3D Object Detection with Complex-YOLO

Autonomous Perception: 3D Object Detection with Complex-YOLO LiDAR object detect

Thomas Dunlap 2 Feb 18, 2022
Various operations like path tracking, counting, etc by using yolov5

Object-tracing-with-YOLOv5 Various operations like path tracking, counting, etc by using yolov5

Pawan Valluri 5 Nov 28, 2022
Differentiable Prompt Makes Pre-trained Language Models Better Few-shot Learners

DART Implementation for ICLR2022 paper Differentiable Prompt Makes Pre-trained Language Models Better Few-shot Learners. Environment

ZJUNLP 83 Dec 27, 2022
Statsmodels: statistical modeling and econometrics in Python

About statsmodels statsmodels is a Python package that provides a complement to scipy for statistical computations including descriptive statistics an

statsmodels 8.1k Jan 02, 2023
LSTM built using Keras Python package to predict time series steps and sequences. Includes sin wave and stock market data

LSTM Neural Network for Time Series Prediction LSTM built using the Keras Python package to predict time series steps and sequences. Includes sine wav

Jakob Aungiers 4.1k Jan 02, 2023
Pytorch Geometric Tutorials

Pytorch Geometric Tutorials

Antonio Longa 648 Jan 08, 2023
MMGeneration is a powerful toolkit for generative models, based on PyTorch and MMCV.

Documentation: https://mmgeneration.readthedocs.io/ Introduction English | 简体中文 MMGeneration is a powerful toolkit for generative models, especially f

OpenMMLab 1.3k Dec 29, 2022