Convert PyTorch tensor to python list

Question

How do I convert a PyTorch Tensor into a python list?

My current use case is to convert a tensor of size [1, 2048, 1, 1] into a list of 2048 elements.

My tensor has floating point values. Is there a solution which also accounts for int and possibly other data types?

question from:https://stackoverflow.com/questions/53903373/convert-pytorch-tensor-to-python-list

Answer 1

I found Tensor.tolist() which gives the following usage example:

>>> import torch
>>> a = torch.randn(2, 2)
>>> a.tolist()
[[0.012766935862600803, 0.5415473580360413],
 [-0.08909505605697632, 0.7729271650314331]]
>>> a[0,0].tolist()
0.012766935862600803

So, to answer the question, use a.squeeze().tolist() to remove all dimensions of size 1.

Also consider .flatten() if a list of lists is not desired.

---

Before I came across .tolist(), I was using:

list = [element.item() for element in tensor.flatten()]

This flattens the tensor into a single dimension then calls .item() to convert each element into a Python number.