Closed jankrepl closed 4 years ago
Write a torch.utils.data.Dataset subclass such that it randomly subsamples both samples and assets. Seems like cvxpylayers does not scale well
torch.utils.data.Dataset
Can be done via DataLoader ..additionally one can cutoff lookback randomly and maybe horizon too
Write a
torch.utils.data.Dataset
subclass such that it randomly subsamples both samples and assets. Seems like cvxpylayers does not scale well