Question about dialogue action prediction dataset

[yukyunglee]

Hi, Thanks for sharing great work!

I have a question about the dataset in the evaluation process. Currently, I want to use dstc2 data to reproduce dialogue action prediction.

However, there are no other instructions for downloading dstc2 or sim_joint, and I downloaded dstc2 and sim_joint from google drive links in TOD-BERT repo.

But in the evaluation code, DATA_DIR below should be satisfied.

for idx in range(args.num_runs):
  DATA_DIR = os.path.join(args.data_dir, str(int(args.data_ratio)))
  DATA_DIR = os.path.join(DATA_DIR, str(idx))

Can you give me more details on the data directory structure to reproduce the results? thank you 😀