iOS does not support all types in model spec about playtorch HOT 5 CLOSED

Thanks for reporting this and for submitting PR #27 to report the differences in the API docs, @mdwelsh!

from playtorch.

So instead of using tuple and list, how can construct my "unpack" part for the tuple of the lists?

from playtorch.

Quick question: how was the IOS object detection demo implemented if tuple wasn't supported on IOS

from playtorch.

@mdwelsh, @JohnZcp, and @kyleliang919, the react-native-pytorch-core release candidate v0.2.0-rc.2 has a PyTorch-inspired API to support a more flexible data processing.

Tuples are now supported on both platforms Android and iOS with the PyTorch Live JavaScript API. For example, the following will run inference and destruct the output (tuple) into y and z.

const x = torch.rand([1, 3, 600, 600]);
const output = await model.forward<Tensor, Tensor[]>(x);
const y = output[0];
const z = output[1];
// or const [y, z] = output;

Check out this example for YOLOv5s integration with PyTorch Live ObjectDetection.ts!

More details on available ops are is in the API docs:

• https://pytorch.org/live/docs/next/api/core/interfaces/torchlive_torch.torch/
• https://pytorch.org/live/docs/next/api/core/interfaces/torchlive_torch.tensor/

Feel free to reach out on Discord or here if any ops are missing for your use cases!

from playtorch.

Closing this issue out since the packer / unpacker approach was removed with 0.2.2. Check out the PyTorch-inspired API in the README as an alternative that avoids these issues: https://github.com/facebookresearch/playtorch/tree/main/react-native-pytorch-core#example-usage

from playtorch.