Provide API to get runtime Tensor buffer, dtype, and shape.

[LPanosTT]

For the frontends to properly reconstruct a tensor during intermediate callback, the data type, shape, and data buffer must be provided. Currently, we can only get the tensor data as a std::vector<float> or as a tt::runtime::Tensor, which itself holds a void pointer to a ttnn::Tensor, which can't be used unless the frontend depends on tt-metal as well.

I believe if the tt::runtime::Tensor held a TensorDesc for itself, or had the shape and dtype of itself then calling getOpOutputTensor should suffice.