junpeiz
junpeiz
I can reproduce it in version 0.16.1. In my case, the broken is also introduced by `tensorflow-macos==2.10.0`. Here are some combinations I tried: - `tensorflow-macos==2.8.0` + `tensorflow-addons==0.16.1`: No issue -...
When I ran the code snippet in https://github.com/apple/coremltools/issues/1423, I met the same error about `baddbmm` not implemented.
This issue also occurs during converting HuggingFace AutoModelForSequenceClassification.
@RahulBhalley Could you provide a code snippet to reproduce the issue? We tested several models in unit test that the converted FFT model is ANE compatible. Your code snippet could...
@hzphzp the PyTorch's `rfft2` is just a special case of `rfftn` (see https://pytorch.org/docs/stable/generated/torch.fft.rfft2.html), and it should be trivial to use `rfftn` to replace the `rfft2` in your model. If you...
@hzphzp from the error message I cannot tell. Could you open another issue with a code snippet to reproduce this error?
https://github.com/apple/coremltools/pull/1936 is an example about using `mb.random_uniform` to implement `torch.rand` op. It should be similar for the `normal_` op by using `mb.random_normal` as suggested in https://github.com/apple/coremltools/issues/1528#issuecomment-1155737497. Feel free to try...
We are working on it. Those complex number related ops (torch.complex, torch.fft.XXX) would be supported in the next release if everything goes well.
Complex Number ops are supported in coremltools 6.2. For the ops mentioned in this issue, `torch.complex` is already supported, and `torch.view_as_real` could be easily changed to stacking the `torch.real` and...
`view_as_real` is supported: https://github.com/apple/coremltools/pull/1944