KnowEE

Code for EMNLP2023 “Multi-Source Multi-Type Knowledge Exploration and Exploitation for Dialogue Generation”

1. Environment

• CUDA >= 11.6
• torch >= 2.0.0

pip install -r requirements.txt

2. Data

data/preprocessed_data

3.Models

download Flan-T5-XXL, ChatGLM, GPT-NEOX from huggingface and change the model path in src/load_models_and_datasets.py

4. Run

sh src/run.sh

If you are interested in our work, please cite:

@inproceedings{conf/emnlp/NiDRL23,
  author       = {Xuanfan Ni and
                  Hongliang Dai and
                  Zhaochun Ren and
                  Piji Li},
  title        = {Multi-Source Multi-Type Knowledge Exploration and Exploitation for
                  Dialogue Generation},
  booktitle    = {Proceedings of the 2023 Conference on Empirical Methods in Natural
                  Language Processing, {EMNLP} 2023, Singapore, December 6-10, 2023},
  year         = {2023},
  url          = {https://aclanthology.org/2023.emnlp-main.771},
}