Prompt-Based Editing for Text Style Transfer

Installation

pip3 install -U pip
conda create -n plm python=3.8
conda activate plm
conda install pytorch torchvision torchaudio pytorch-cuda=11.8 -c pytorch -c nvidia
conda install transformers
pip3 install -r requirements.txt

Model downloading

We use Eleuther AI's gpt-j-6B as the main architecture.

Please also download the following models for experiments.

gpt2

RoBERTa-Large

distilbert-base-uncased-finetuned-sst-2-english

Model running

bash run.sh

Evaluation

Put the generated outputs in 0-1.txt into output.txt, then the generated outputs 1-0.txt into output.txt.

For Accuracy

cd evaluator/
python3 sentiment.py --gen_path ../data/yelp/output.txt 

For BLEU

perl multi-bleu.perl ../data/yelp/references/ref0 ../data/yelp/references/ref1 ../data/yelp/references/ref2 ../data/yelp/references/ref3 < ../data/yelp/output.txt 

Cite our work

If you find this repo helpful, please consider citing our work:

@inproceedings{luo-etal-2023-prompt,
    title = "Prompt-Based Editing for Text Style Transfer",
    author = "Luo, Guoqing  and
      Han, Yu  and
      Mou, Lili  and
      Firdaus, Mauajama",
    editor = "Bouamor, Houda  and
      Pino, Juan  and
      Bali, Kalika",
    book title = "Findings of the Association for Computational Linguistics: EMNLP 2023",
    year = "2023",
    url = "https://aclanthology.org/2023.findings-emnlp.381",
    pages = "5740--5750",
}