Black-Box-Tuning

Source code for paper "Black-Box Tuning for Language-Model-as-a-Service".

Being busy recently, the code in this repo and this tutorial will be very brief. Please let me know if you find any issues.

Prepare your environment

The implementation of Black-Box Tuning is quite simple, you can check our code and easily implement it in your own environment. Or you can create a new environment to run our implementation, which is based on Nevergrad, Transformers and FastNLP. Optionally, we use fitlog to monitor experimental results. You can uncomment the fitlog-related lines in our code to use it.

conda create --name bbt python=3.8
conda activate bbt
pip install transformers==4.1.1
pip install datasets
pip install fastNLP
pip install nevergrad
pip install sklearn
git clone https://github.com/txsun1997/Black-Box-Tuning
cd Black-Box-Tuning

Optimize your prompt without gradients

Now you can run Black-Box Tuning with run.sh:

bash run.sh

Results will be saved in a directory named results/. In general, you will obtain the following results:

To reproduce other experiments in our paper, change the arguments of bbt.py, for example,

python bbt.py --task_name "agnews" --n_prompt_tokens 50 --intrinsic_dim 500 --k_shot 16 --device "cuda:0" --seed 42 --loss_type "hinge" --cat_or_add "add" --budget 8000

Cite

If you find this work helpful, please cite:

@article{sun2022bbt,
  title={Black-Box Tuning for Language-Model-as-as-Service}, 
  author={Tianxiang Sun and Yunfan Shao and Hong Qian and Xuanjing Huang and Xipeng Qiu},
  journal={arXiv preprint arXiv:2201.03514},
  year={2022}
}