TensorDict

[GeorgeS2019]

TensorDict is a pytorch dedicated tensor container.

https://github.com/pytorch-labs/tensordict

TensorDict is a dictionary-like class that inherits properties from tensors, such as indexing, shape operations, casting to device or point-to-point communication in distributed settings.

The main purpose of TensorDict is to make code-bases more readable and modular by abstracting away tailored operations:

==> Question?
Should TensorDict be implemented as same additional TorchSharp features in similar way as TorchAudio and TorchVision?

[GeorgeS2019]

Related discussion
[WIP] Compare and understand TorchSharp Memory Management through PyTorch's Approach