https://github.com/amazon-science/musketeer

Musketeer: A Vision-Language Model with Task Explanation Prompts

https://github.com/amazon-science/musketeer

Science Score: 10.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • CITATION.cff file
  • codemeta.json file
  • .zenodo.json file
  • DOI references
  • Academic publication links
    Links to: arxiv.org
  • Committers with academic emails
  • Institutional organization owner
  • JOSS paper metadata
  • Scientific vocabulary similarity
    Low similarity (10.2%) to scientific vocabulary
Last synced: 11 months ago · JSON representation

Repository

Musketeer: A Vision-Language Model with Task Explanation Prompts

Basic Info
  • Host: GitHub
  • Owner: amazon-science
  • License: apache-2.0
  • Language: Python
  • Default Branch: main
  • Homepage:
  • Size: 3.62 MB
Statistics
  • Stars: 0
  • Watchers: 5
  • Forks: 0
  • Open Issues: 0
  • Releases: 0
Created about 3 years ago · Last pushed over 2 years ago
Metadata Files
Readme Contributing License Code of conduct

README.md

Introduction

This codebase is an official implementation for Musketeer, which aims at building a sequence-to-sequence vision-language model whose parameters are jointly trained on all tasks (all for one) and fully shared among multiple tasks (one for all).

Requirements

  • python 3.7.4
  • pytorch 1.8.1
  • torchvision 0.9.1
  • JAVA 1.8 (for COCO evaluation)

Installation

This implementation based on OFA and fairseq.

bash pip install -r requirements.txt git clone https://github.com/pytorch/fairseq cd fairseq pip install --editable ./

Datasets and Pretrained Checkpoints

Please prepare data and OFA-pretrained checkpoints according to datasets.md and checkpoints.md.

After download and unzip these datasets in your data directory, make sure that the directory structure is arranged as follows,

├── your_data_directory
│   ├── caption_data
│   ├── snli_ve_data
│   ├── refcoco_data
│   ├── vqa_data
│   ├── coco
│   ├── imagenet_1k_data
│   ├── gigaword

then run bash export DATADIR=your_data_directory

For preparing pretrained checkpoints, please run bash mkdir checkpoints wget https://ofa-beijing.oss-cn-beijing.aliyuncs.com/checkpoints/ofa_base.pt wget https://ofa-beijing.oss-cn-beijing.aliyuncs.com/checkpoints/ofa_large.pt mv ofa_base.pt ofa_large.pt checkpoints

Training

To train Musketeer with Task Explanation Prompt (TEP), please run bash cd run_scripts/musketeer bash train_musketeer.sh To train Musketeer with Base Prompt (baseP), please run bash cd run_scripts/musketeer bash train_musketeer_baseP.sh

Evaluation on Visual Grounding

bash cd run_scripts/vg bash evaluate_refcoco_base.sh your_checkpoint_file replace your_checkpoint_file with your trained model file dir.

For evaluating other tasks, please use the scripts in run_scripts.

Related Codebase

We thanks following (but not limited to) researchers for sharing their code, * OFA * fairseq

Acknowledgement

This code was developed by Zhaoyang Zhang while he was interning at the AWS Rekognition Team.

Citation

If this code helps your research or project, please cite

@article{zhang2023musketeer, title={Musketeer: Joint Training for Multi-task Vision Language Model with Task Explanation Prompts}, author={Zhang, Zhaoyang and Shen, Yantao and Shi, Kunyu and Cai, Zhaowei and Fang, Jun and Deng, Siqi and Yang, Hao and Modolo, Davide and Tu, Zhuowen and Soatto, Stefano}, journal={arXiv preprint arXiv:2305.07019}, year={2023} }

Contact Info

If you have any question, feel free to contact Zhaoyang Zhang or his mentor at AWS, Yantao Shen

Zhaoyang Zhang: zhaoyangzhang@link.cuhk.edu.hk Yantao Shen: yantaos@amazon.com

Owner

  • Name: Amazon Science
  • Login: amazon-science
  • Kind: organization

GitHub Events

Total
  • Watch event: 1
  • Pull request event: 1
Last Year
  • Watch event: 1
  • Pull request event: 1

Committers

Last synced: about 1 year ago

All Time
  • Total Commits: 5
  • Total Committers: 3
  • Avg Commits per committer: 1.667
  • Development Distribution Score (DDS): 0.4
Past Year
  • Commits: 0
  • Committers: 0
  • Avg Commits per committer: 0.0
  • Development Distribution Score (DDS): 0.0
Top Committers
Name Email Commits
Yantao Shen y****s@a****m 3
Yantao Shen 4****n 1
Amazon GitHub Automation 5****o 1
Committer Domains (Top 20 + Academic)

Issues and Pull Requests

Last synced: about 1 year ago

All Time
  • Total issues: 0
  • Total pull requests: 0
  • Average time to close issues: N/A
  • Average time to close pull requests: N/A
  • Total issue authors: 0
  • Total pull request authors: 0
  • Average comments per issue: 0
  • Average comments per pull request: 0
  • Merged pull requests: 0
  • Bot issues: 0
  • Bot pull requests: 0
Past Year
  • Issues: 0
  • Pull requests: 0
  • Average time to close issues: N/A
  • Average time to close pull requests: N/A
  • Issue authors: 0
  • Pull request authors: 0
  • Average comments per issue: 0
  • Average comments per pull request: 0
  • Merged pull requests: 0
  • Bot issues: 0
  • Bot pull requests: 0
Top Authors
Issue Authors
Pull Request Authors
  • dependabot[bot] (1)
Top Labels
Issue Labels
Pull Request Labels
dependencies (1) python (1)