How to train on my own dataset?

[young169]

The dataset returns a list of str, so error happens at train_model.py line 79-line81
"
labels = Variable(torch.from_numpy(np.array(labels)).cuda())
labels_swap = Variable(torch.from_numpy(np.array(labels_swap)).cuda())
swap_law = Variable(torch.from_numpy(np.array(swap_law)).float().cuda())
"
cannot convert labels to np.ndarray