Learning to compose soft prompts for compositional zero-shot learning.

Last update: Jan 02, 2023

Related tags

Overview

Compositional Soft Prompting (CSP)

Compositional soft prompting (CSP), a parameter-efficient learning technique to improve the zero-shot compositionality of large-scale pretrained vision-language models (VLMs) without the overhead of fine-tuning the entire model.

Reference Paper: Learning to Compose Soft Prompts for Compositional Zero-Shot Learning

If you find CSP helpful, please cite our paper:

@article{csp2022,
  author = {Nayak, Nihal V. and Yu, Peilin and Bach, Stephen H.},
  title = {Learning to Compose Soft Prompts for Compositional Zero-Shot Learning},
  volume = {arXiv:2204.03574 [cs.LG]},
  year = {2022},
}

Setup

conda create --name clip python=3.7
conda activate clip
pip3 install torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu113
pip3 install ftfy regex tqdm scipy pandas
pip3 install git+https://github.com/openai/CLIP.git

Alternatively, you can use pip install -r requirements.txt to install all the dependencies.

Download Dataset

We experiment with three datasets: MIT-States, UT-Zappos, and C-GQA.

sh download_data.sh

If you already have setup the datasets, you can use symlink and ensure the following paths exist: data/<dataset> where <datasets> = {'mit-states', 'ut-zappos', 'cgqa'}.

Training

python -u train.py \
  --dataset mit-states \
  --model ViT-L/14 \
  --experiment_name csp \
  --seed 0 \
  --epochs 20 \
  --lr 5e-05 \
  --attr_dropout 0.3 \
  --weight_decay 0.00001 \
  --train_batch_size 64 \
  --gradient_accumulation_steps 2 \
  --context_length 8 \
  --save_path data/model/mit-states/sample_model \
  --save_every_n 1

You can replace --dataset with {mit-states, ut-zappos, cgqa}. The best hyperparameters are included in the paper.

Evaluation

We evaluate our models in two settings: closed-world and open-world.

Closed-World Evaluation

python -u evaluate.py \
  --dataset mit-states \
  --soft_embeddings data/model/mit-states/sample_model/soft_embeddings_epoch_20.pt \
  --context_length 16 \
  --text_encoder_batch_size 36 \
  --eval_batch_size 16 \
  --experiment_name csp

Open-World Evaluation

For our open-world evaluation, we compute the feasbility calibration and then evaluate on the dataset.

Feasibility Calibration

We use GloVe embeddings to compute the similarities between objects and attributes. Download the GloVe embeddings in the data directory:

cd data
wget https://nlp.stanford.edu/data/glove.6B.zip

Move glove.6B.300d.txt into data/glove.6B.300d.txt.

To compute feasibility calibration for each dataset, run the following command:

python -u datasets/feasibility.py --dataset mit-states

The feasibility similarities are saved at data/feasibility_<dataset>.pt.

Evaluation

The open-world evaluation with the thresholds (feasibility calibration).

python -u evaluate.py \
  --dataset mit-states \
  --soft_embeddings data/model/mit-states/sample_model/soft_embeddings_epoch_5.pt \
  --context_length 16 \
  --text_encoder_batch_size 36 \
  --eval_batch_size 256 \
  --experiment_name czsl \
  --threshold <threshold> \
  --open_world

If <threshold> is None, then the model picks the best threshold on the validation set. We use the following thresholds:

Dataset	Threshold
mit-states	0.4069159426
ut-zappos	0.5299109123
cgqa	0.49937106273612186

Note: We use 256GB of cpu memory to evaluate cgqa.

Generalization to Higher-Order Compositions

Evaluate the trained CSP vocabulary on the new AAO-MIT-States dataset.

python aao/evaluate_att_att_obj.py \
  --experiment_name csp \
  --soft_embeddings data/model/mit-states/sample_model/soft_embeddings_epoch_20.pt

We thank Andrew Delworth and Elise Carman for helping us annotate this dataset.

Generalization to Mixed Pretrained and Fine-Tuned Vocabulary

Ablation experiment to train and evaluate CSP with reduced fine-tuned vocabulary. We run experiment on the ut-zappos dataset.

Training

python -u mix/mix_train.py \
  --dataset ut-zappos \
  --model ViT-L/14 \
  --experiment_name mix_csp \
  --seed 0 \
  --epochs 20 \
  --lr 5e-04 \
  --attr_dropout 0.2 \
  --weight_decay 0.00001 \
  --train_batch_size 64 \
  --context_length 8 \
  --save_path data/model/ut-zappos/mix_train_model_0.25 \
  --save_every_n 5 \
  --attr_keep_ratio 0.25 \
  --gradient_accumulation_steps 2

We change the --attr_keep_ratio to {0.25, 0.50, 0.75}.

Evaluation

python -u mix/evaluate_mix_train.py \
  --dataset ut-zappos \
  --soft_embeddings data/model/ut-zappos/mix_train_model_0.25/soft_embeddings.pt \
  --context_length 16 \
  --text_encoder_batch_size 36 \
  --eval_batch_size 256 \
  --experiment_name csp

Credits

The project uses openly available model, code, and datasets. Please see the credits.

Learning to compose soft prompts for compositional zero-shot learning.

Related tags

Overview

Compositional Soft Prompting (CSP)

Setup

Download Dataset

Training

Evaluation

Closed-World Evaluation

Open-World Evaluation

Feasibility Calibration

Evaluation

Generalization to Higher-Order Compositions

Generalization to Mixed Pretrained and Fine-Tuned Vocabulary

Training

Evaluation

Credits

Owner

Bats Research

pwncat module that automatically exploits CVE-2021-4034 (pwnkit)

Python program that generates secure passwords.

HashDB API hash lookup plugin for IDA Pro

Open-source keylogger write in python

Automatic ProxyShell Exploit

The next level Python obfuscator, nearly impossible to deobfuscate.

A gui application used for network reconnaissance while pentesting

Tool ini berfungsi untuk membuat virus secara instan

S2-061 的payload，以及对应简单的PoC/Exp

Natas teaches the basics of serverside web-security.

Log4j exploit catcher, detect Log4Shell exploits and try to get payloads.

In this program, I generate the strong password from existing password without class using python. The special thing in this program is that is generate very strong password from our existing password. It replaces some elements from password and put another elements in it.

CloakifyFactory & the Cloakify Toolset - Data Exfiltration & Infiltration In Plain Sight;

FBGen is simple facebook user based wordlist generator using Username/ID and cookie.

This repository consists of the python scripts for execution and automation of vivid tasks.

The probability of having the password you want in the PassMaker is +90%!!

Tinyman exploit finder - Tinyman exploit finder for python

This is a js front-end encryption blasting account and password tools

Discord Token Stealer Malware Protection

DoSer.py - Simple DoSer in Python