ExPred

1. This is the implementation of the paper Explain and Predict, and then Predict Again (accepted in WSDM2021).
2. This code is implemented based on the pipeline model of the Eraserbenchmark. All data used by the model can be found from the Eraser Benchmark, too.
3. I just perform the experiments on this code using the different hyperparameters and reproduce the results.

Usage:

1. Install the required packages from the enviroment.yml using the command conda env create -f environment.yml , after that activate the enviroment.
2. Change "--data_dir" to "data/{movies/fever/multirc}" and "--conf" to "params/{movies,fever,multirc}_expred.json" .
3. Run the file train.py using python train.py .

**Note: **

1. Depending on your hardware you may have to change the batch_size in the config file.
2. For fever dataset , you have to use a scheduler. So, to use it uncomment line no 104 and 105 in mtl_evidence_classifier.py .