Skip to main content

PromptCraft: A Prompt Perturbation Toolkit for Prompt Robustness Analysis

Project description

PromptCraft

A Prompt Perturbation Toolkit for Prompt Robustness Analysis

Code License License Python 3.9+

Table of Contents

Installation

pip install promptcraft

Character Editing

from promptcraft import character

sentence = "Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May."
level = 0.25  # Percentage of characters that will be edited
character_tool = character.CharacterPerturb(sentence=sentence, level=level)

Character Replacement

char_replace = character_tool.character_replacement()

Character Deletion

char_delete = character_tool.character_deletion()

Character Insertion

char_insert = character_tool.character_insertion()

Character Swap

char_swap = character_tool.character_swap()

Keyboard Typos

char_keyboard = character_tool.keyboard_typos()

Optical Character Recognition

char_ocr = character_tool.optical_character_recognition()

Word Manipulation

from promptcraft import word

sentence = "Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May."
level = 0.25  # Percentage of words that will be manipulated
word_tool = word.WordPerturb(sentence=sentence, level=level)

Synonym Replacement

word_synonym = word_tool.synonym_replacement()

Word Insertion

word_insert = word_tool.word_insertion()

Word Swap

word_swap = word_tool.word_swap()

Word Deletion

word_delete = word_tool.word_deletion()

Insert Punctuation

word_punctuation = word_tool.insert_punctuation()

Word Split

word_split = word_tool.word_split()

Sentence Paraphrasing

from promptcraft import sentence

sen = "Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May."
sentence_tool = sentence.SentencePerturb(sentence=sen)

Back Translation by Hugging Face

back_trans_hf = sentence_tool.back_translation_hugging_face()

Back Translation by Google Translator

back_trans_google = sentence_tool.back_translation_google()

Paraphrasing

sen_paraphrase = sentence_tool.paraphrase()

Formal Style

sen_formal = sentence_tool.formal()

Casual Style

sen_casual = sentence_tool.casual()

Passive Style

sen_passive = sentence_tool.passive()

Active Style

sen_active = sentence_tool.active()

Parallel Processing

Since all the methods are executed on the CPU, they can be performed in parallel using the multiprocessing package.

Structure of the Code

At the root of the project, you will see:

.
├── LICENSE
├── README.md
├── promptcraft
│   ├── __init__.py
│   ├── character.py
│   ├── parrot.py
│   ├── sentence.py
│   ├── styleformer.py
│   └── word.py
└── setup.py

Acknowledgement

This work was finished during my 2023 fall semester research rotation at the Dependable Computing Laboratory, Department of Electrical and Computer Engineering, Boston University, under the supervision of Prof. Wenchao Li.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

promptcraft-0.3.2.tar.gz (14.7 kB view details)

Uploaded Source

Built Distribution

promptcraft-0.3.2-py3-none-any.whl (15.1 kB view details)

Uploaded Python 3

File details

Details for the file promptcraft-0.3.2.tar.gz.

File metadata

  • Download URL: promptcraft-0.3.2.tar.gz
  • Upload date:
  • Size: 14.7 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.2 CPython/3.10.9

File hashes

Hashes for promptcraft-0.3.2.tar.gz
Algorithm Hash digest
SHA256 acdc17049167eef4e8c5b1172095b13cfc3a470cb0403f5cf6472e7adb51fc10
MD5 b26c17af3ca9a7144fb86b4b8a7f73eb
BLAKE2b-256 d73bf89f333b2a6d5d1067ab411fcc98079e30f849e93cba36a0cabb745be695

See more details on using hashes here.

File details

Details for the file promptcraft-0.3.2-py3-none-any.whl.

File metadata

  • Download URL: promptcraft-0.3.2-py3-none-any.whl
  • Upload date:
  • Size: 15.1 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.2 CPython/3.10.9

File hashes

Hashes for promptcraft-0.3.2-py3-none-any.whl
Algorithm Hash digest
SHA256 b9b99e8d5a4ffaf6ccffc7a0b030555cf3fa22d34e30b3301fda52e3f2fcb542
MD5 50aab74fe214704b2154db8e30f51ecb
BLAKE2b-256 bf24a3becb8f587470bf5810f917f5b53ecc20f423300c8a6045842c70c21113

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page