[FIXED] Read multiple .gz file and return it in one tensor

Issue

I am trying to read multiple .gz file and return its content in one tensor as follows:

with ReadHelper('ark: gunzip -c /home/mnabih/kaldi/egs/timit/s5/exp/mono_ali/*.gz|') as reader:
    for key, b in reader:
        #print(type(b))
        c = torch.from_numpy(b)
        labels = torch.cat(c)

Unfortunately, it gives me this error:

cat(): argument 'tensors' (position 1) must be tuple of Tensors, not Tensor

Solution

As the error message explains, c is a tensor. To use torch.cat() you must pass a group of tensors or a list. To solve your problem you may use:

temp = list()
for key, b in reader:
    temp.append(torch.from_numpy(b))
labels = torch.cat(temp)

For more, you can check the manual here

Answered By - André Pacheco