I’m following the instructions that say:

dataset_type = "train"
dataset = FIDataset(dataset_type)

to train the model on the train dataset. However, the code for FIDataset actually requires:

FIDataset(
    dataset_type: str,
    dataset_path: str | Path,
    ...
)

so if I only pass dataset_type, I get a TypeError because dataset_path is missing.

Additionally, my BenchmarkDatasets folder does not contain any file named data_train.npy. Instead, I see files like Train_Dst_NoAuction_ZScore_CF_7.txt and so on. The FIDataset script specifically looks for a file named data_train.npy (or data_test.npy, etc.).

FileNotFoundError: [Errno 2] No such file or directory: 'BenchmarkDatasets/data_train.npy'

How I can handle the missing data_train.npy? Is there a step I’m missing to generate them?

Any guidance or suggestions would be a huge help. Thank you!