This repo is the official code of our work Variety Is the Spice of Life: Detecting Fake News with Dynamic Contextual Representations
This code acknowledges and follows to Generalizing to the Future: Mitigating Entity Bias in Fake News Detection
cudatoolkit==11.8.0
torch==1.13.1
numpy==1.23.5
- Download datasets from GITHUB LINK, and place them to
./FENDER_en/data
and./FENDER_ch/data
, respectively; - Run the scripts
./FENDER_en/run.sh
or./FENDER_ch/run.sh
to rerun the experiments, or you can directly run
python main.py --para_len 32 --seed 1000
The log files are saved in ./logs/param
, and the checkpoints can be loaded from ./para_model
.
We acknowledge the source code provided by GITHUB LINK.
NONE