PyTorch 1.8.1 Compatibility with Training and Evaluation

[atagade]

Expanding on the PR's raised by #12 and #10, this PR focuses on providing compatibility for PyTorch 1.8.1 for both evaluation and training purposes. With this PR it is possible to train agents using different label mappings and feature maps.

[Throne-myworld]

Expanding on the PR's raised by #12 and #10, this PR focuses on providing compatibility for PyTorch 1.8.1 for both evaluation and training purposes. With this PR it is possible to train agents using different label mappings and feature maps.
But when I running the code on win10 and using the latest python,pytorch version is 1.8.1,I got this
quesiton:
File "D:\PyCharmproject\Arnold-master\Arnold-master\src\model\dqn\base.py", line 81, in base_forward
output = torch.cat([conv_output] + embeddings, dim=1)
RuntimeError: torch.cat(): Tensors must have same number of dimensions: got 2 and 3
which is quiet confusing for me to solve it

[atagade]

Could you tell me the command you used to run training?