Im getting started with the C++ frontend, I wanted to ask how I could create a dataset from two tensors, one with labels and another with features
The torch::data::datasets::TensorDataset
, which is the equivalent of torch.utils.data.TensorDataset
in Python, might work.
Best regards
Thomas
Thanks, I have exactly same problem with how to use TensorDataset,
The python code is
return TensorDataset(torch.FloatTensor(np.array(data), torch.LongTensor(np.array(labels)))
May I ask how to convert the above in C++ please ?
I write torch::data::datasets::TensorDataset T1[]{};
How to send data and label to T1 in this case ?
Thanks.