prompt2slip
This library is testing the ethics of language models by using natural adversarial texts.
This tool allows for short and simple code and validation with little effort.
Extensibility to be applied to any language model in any problem setting by inheriting from the base class.
Install
pip install prompt2slip
Usage
The simplest sample looks like this.
import prompt2slip
from transformers import GPT2Tokenizer, GPT2Model
tokenizer = GPT2Tokenizer.from_pretrained('gpt2')
model = GPT2Model.from_pretrained('gpt2')
base_text = ["This project respects pytorch developers."]
target_word = ["keras"]
target_ids = torch.Tensor(tokenizer.convert_tokens_to_ids(target_words))
attaker = CLMAttacker(model,tokenizer)
output = attacker.attack_by_text(texts,target_ids)
More realistic use cases are stored in examples.
What It Dose
"prompt2slip" provides the function to search for prompts which cause appearance of any specific word against a pre trained natural language generation model. Furthermore, with user customization, it can be applied to a wide range of tasks, including classification tasks.If you want to generate a hostile sample for a classification model, you can simply override the method to compute the adversarial loss function to generate a natural adversarial text.
The unique feature of this library is that it can generate test cases for verifying the danger of a pre-trained natural language model with a few lines of code.
References
More About This Project
License
Development
Install Package
Install poetry
poetry install
Test
Running tests with pytest.
poetry run pytest --cov .
1.2k Dec 18, 2022
125 Dec 20, 2022
2 Jan 17, 2022
1.2k Dec 21, 2022
2 May 17, 2022
2.9k Jan 2, 2023
65 Sep 13, 2022
1.6k Jan 05, 2023
15 Oct 24, 2022
1 Jan 06, 2022
544 Jan 09, 2023
170 Dec 13, 2022
16 Dec 11, 2022
1.6k Dec 27, 2022
207 Dec 25, 2022
3 Feb 18, 2022
13 Jun 09, 2022
290 Dec 20, 2022
902 Jan 06, 2023
12 Nov 05, 2022
73 Dec 28, 2022
217 Dec 05, 2022
1.1k Dec 27, 2022
67 Nov 14, 2022
1 Oct 15, 2021
4 Nov 24, 2021
74 Dec 13, 2022