Toby Roseman

`is44` is the name of the symbol. It doesn't mean `44`. Are you using flexible shape when you call `ct.convert`?

`is44` just mean it's the 44th symbol created during the conversion. It doesn't tell you anything about its' values. In fact the value is not known at conversion time. Have...

> I also commented out [the two checks](https://github.com/apple/coremltools/blob/40f6705a6a4c5ef1a616f01b444f1d6e5a0c1f59/coremltools/converters/mil/mil/ops/defs/iOS15/conv.py#L163C11-L168) in question and the model exported fine. Are you also able to get correct predictions from that exported Core ML model?

We have an unmerged pull request that adds support for `roi_align` (not `torchvision::roi_align`): #1509.

This error looks unrelated to your previous error. I think it's an entirely separate issue.

We do have support for `im2col`: https://github.com/apple/coremltools/blob/ea1d2deffd52f18e75962e2e600a4c29c1bab2f5/coremltools/converters/mil/frontend/torch/ops.py#L8288 This seems to be a less general form of `torch.unfold`: https://github.com/apple/coremltools/blob/ea1d2deffd52f18e75962e2e600a4c29c1bab2f5/coremltools/converters/mil/frontend/torch/ops.py#L8292

Thanks for the minimal example. I can reproduce the issue with it.

Take a look at our [Model Debugging User Guide Page](https://apple.github.io/coremltools/docs-guides/source/mlmodel-debugging-perf-utilities.html).

Please give us some more details here. How big is your model? What are you trying to do? What code are you running to do that? How long is it...

Thanks @metascroy for the pull request. Can you add a unit test?