ifsheldon

@rahulbhalley You may want to take a look at [the TF release by Apple](https://github.com/apple/tensorflow_macos), but Apple did not release the source code, which is not so friendly

Sorry for late reply. And yes, this can be helpful but I didn't do it because I didn't need this exact feature in my dogfooding. However, for `Tube`, I have...

If I understand correctly, you need to relate the shape of output with the shapes of inputs. `Tube` manages Taichi field creation automatically on your behalf, so you need to...

> Does it mean I could not register "input" as an output tensor, but rather register it as an input tensor and thus not mind about shapes? I think you...

I wrote a prototype in the branch `dim_calc`. Maybe you can help test it. You'll need to write a subclass of `DimensionCalculator`, which needs only one method implemented. The arguments...

related Taichi [issue](https://github.com/taichi-dev/taichi/issues/5282).

The Taichi issue is closed. Looking forward to Taichi new release!

The new release of Taichi 1.1.3 includes the PR for performance improvement, so this issue is resolved.

Reopen this because Taichi still has the problem on other backends (Vulkan and Metal) while the problem was fixed on CPU and CUDA backends. See this [issue](https://github.com/taichi-dev/taichi/issues/4068).

`check_field_needs_grad` throws an error due to this [issue](https://github.com/taichi-dev/taichi/issues/5258). Waiting for upstream fix. --- Fixed in v0.6.3