Code for our SIGIR 2022 accepted paper : P3 Ranker: Mitigating the Gaps between Pre-training and Ranking Fine-tuning with Prompt-based Learning and Pre-finetuning

Related tags

Documentation P3Ranker
Overview

P3 Ranker

Implementation for our SIGIR2022 accepted paper:

P3 Ranker: Mitigating the Gaps between Pre-training and Ranking Fine-tuning with Prompt-based Learning and Pre-finetuning

Project Structures

├── commands
│   ├── bert.sh
│   ├── p3ranker.sh
│   ├── prop_ft.sh
│   ├── roberta.sh
│   └── t5v11.sh
├── Prefinetune
│   ├── mnli_dataloader.py
│   ├── mnli_dataset.py
│   ├── mnli_model.py
│   ├── README.md
│   ├── train_mnli.sh
│   ├── train_nq.sh
│   ├── train.py
│   └── utils.py
├── src
│   ├── data
│   │    ├── datasets
│   │    │   ├── __init__.py
│   │    │   ├── bert_dataset.py
│   │    │   ├── bertmaxp_dataset.py
│   │    │   ├── dataset.py
│   │    │   ├── edrm_dataset.py
│   │    │   ├── roberta_dataset.py
│   │    │   └── t5_dataset.py
│   │    └── tokenizers
│   │        ├── __init__.py
│   │        ├── tokenizer.py
│   │        └── word_tokenizer.py
│   ├── extractors
│   │    ├── __init__.py
│   │    └── classic_extractor.py
│   ├── metrics
│   │    ├── __init__.py
│   │    └── metric.py
│   ├── models
│   │    ├── __init__.py
│   │    ├── bert_maxp.py
│   │    ├── bert_prompt_.py
│   │    ├── bert.py
│   │    ├── conv_knrm.py
│   │    ├── edrm.py
│   │    ├── knrm.py
│   │    ├── t5.py
│   │    └── tk.py
│   ├── modules
│   │    ├── attentons
│   │    │   ├── __init__.py
│   │    │   ├── multi_head_attention.py
│   │    │   └── scaled_dot_product_attention.py
│   │    ├── embedders
│   │    │   ├── __init__.py
│   │    │   └── embedder.py
│   │    ├── encoders
│   │    │   ├── __init__.py
│   │    │   ├── cnn_encoder.py
│   │    │   ├── feed_forward_encoder.py
│   │    │   ├── positional_encoder.py
│   │    │   └── transformer_encoder.py
│   │    └── matchers
│   │        ├── __init__.py
│   │        └── kernel_matcher.py
│   ├── __init__.py
│   └── utils.py
├── README.md
├── requirements.txt
├── train.py
└── utils.py

Prerequisites

Install dependencies:

git clone https://github.com/NEUIR/P3Ranker.git
cd P3-Rankers
pip install -r requirements.txt

Data Preparation

We will release our few-shot dataset soon.

Prompt Generation

Details about the Discrete Prompt Generation can be find in https://github.com/princeton-nlp/LM-BFF and our paper

Prefinetune

cd Reproduce

And you will find how to do prefinetune.

Reproduce our results

Directly run the scripts we stored in './commands' can reproduce our results. One example is shown below:

bash commands/bert.sh 5

The above command is for reproducing results in our 5-q few-shot scenarios mentioned in our paper.

Contact

Please send email to [email protected].

You might also like...
A collection and example code of every topic you need to know about in the basics of Python.
A collection and example code of every topic you need to know about in the basics of Python.

The Python Beginners Guide: Master The Python Basics Tonight This guide is a collection of every topic you need to know about in the basics of Python.

Ahmed Baari 1 Dec 19, 2021
Show Rubygems description and annotate your code right from Sublime Text.

Gem Description for Sublime Text Show Rubygems description and annotate your code. Just mouse over your Gemfile's gem definitions to show the popup. s

Nando Vieira 2 Dec 19, 2022
Some code that takes a pipe-separated input and converts that into a table!

tablemaker A program that takes an input: a | b | c # With comments as well. e | f | g h | i |jk And converts it to a table: ┌───┬───┬────┐ │ a │ b │

CodingSoda 2 Aug 30, 2022
Some of the best ways and practices of doing code in Python!

Pythonicness ❤ This repository contains some of the best ways and practices of doing code in Python! Features Properly formatted codes (PEP 8) for bet

Samyak Jain 2 Jan 15, 2022
graphical orbitational simulation of solar system planets with real values and physics implemented so you get a nice elliptical orbits. you can change timestamp value or scale from source code idc.
graphical orbitational simulation of solar system planets with real values and physics implemented so you get a nice elliptical orbits. you can change timestamp value or scale from source code idc.

solarSystemOrbitalSimulation graphical orbitational simulation of solar system planets with real values and physics implemented so you get a nice elli

Mega 3 Mar 3, 2022
Run `black` on python code blocks in documentation files

blacken-docs Run black on python code blocks in documentation files. install pip install blacken-docs usage blacken-docs provides a single executable

Anthony Sottile 460 Dec 23, 2022
The source code that powers readthedocs.org

Welcome to Read the Docs Purpose Read the Docs hosts documentation for the open source community. It supports Sphinx docs written with reStructuredTex

Read the Docs 7.4k Dec 25, 2022
Documentation of the QR code found on new Austrian ID cards.

Austrian ID Card QR Code This document aims to be a complete documentation of the format used in the QR area on the back of new Austrian ID cards (Per

Gabriel Huber 9 Dec 12, 2022
Automated generation of real Swagger/OpenAPI 2.0 schemas from Django REST Framework code.
Automated generation of real Swagger/OpenAPI 2.0 schemas from Django REST Framework code.

drf-yasg - Yet another Swagger generator Generate real Swagger/OpenAPI 2.0 specifications from a Django Rest Framework API. Compatible with Django Res

Cristi Vîjdea 3k Dec 31, 2022
Owner
null
GitHub
Layout Parser is a deep learning based tool for document image layout analysis tasks.

A Python Library for Document Layout Understanding

layout-parser 3.4k Dec 30, 2022
Quick tutorial on orchest.io that shows how to build multiple deep learning models on your data with a single line of code using python

Deep AutoViML Pipeline for orchest.io Quickstart Build Deep Learning models with a single line of code: deep_autoviml Deep AutoViML helps you build te

Ram Seshadri 6 Oct 2, 2022
Documentation generator for C++ based on Doxygen and mosra/m.css.

mosra/m.css is a Doxygen-based documentation generator that significantly improves on Doxygen's default output by controlling some of Doxygen's more unruly options, supplying it's own slick HTML+CSS generation and adding a fantastic live search feature.

Mark Gillard 109 Dec 7, 2022
Explicit, strict and automatic project version management based on semantic versioning.

Explicit, strict and automatic project version management based on semantic versioning. Getting started End users Semantic versioning Project version

Dmytro Striletskyi 6 Jan 25, 2022
Dev Centric Tools for Mkdocs Based Documentation

docutools MkDocs Documentation Tools For Developers This repo is providing a set of plugins for mkdocs material compatible documentation. It is meant

Axiros GmbH 14 Sep 10, 2022
Fast syllable estimation library based on pattern matching.

Syllables: A fast syllable estimator for Python Syllables is a fast, simple syllable estimator for Python. It's intended for use in places where speed

ProseGrinder 26 Dec 14, 2022
A document format conversion service based on Pandoc.

reformed Document format conversion service based on Pandoc. Usage The API specification for the Reformed server is as follows: GET /api/v1/formats: L

David Lougheed 3 Jul 18, 2022
A simple document management REST based API for collaboratively interacting with documents

documan_api A simple document management REST based API for collaboratively interacting with documents.

Shahid Yousuf 1 Jan 22, 2022
This repo provides a package to automatically select a random seed based on ancient Chinese Xuanxue

?? Random Luck Deep learning is acturally the alchemy. This repo provides a package to automatically select a random seed based on ancient Chinese Xua

Tong Zhu（朱桐） 33 Jan 3, 2023
script to calculate total GPA out of 4, based on input gpa.csv

gpa_calculator script to calculate total GPA out of 4 based on input gpa.csv to use, create a total.csv file containing only one integer showing the t

Mohamad Bastin 1 Feb 7, 2022