Error on learner.predict() after update

Part 1 (2019)
1

Hello fellows,

I had a perfectly functioning model and then after an update in fastai and pytorch I can’t get it to work.

When i try to call predict(tensor) I got the following error:

IndexError: index 0 is out of bounds for axis 0 with size 0

I’ve already checked the tensor and the dimensions are correct. Is the same tensor that were working

The error is followed by a series of warnings:

SourceChangeWarning: source code of class ‘torchvision.models.densenet._DenseBlock’ has changed. you can retrieve the original source code by accessing the object’s source attribute or set torch.nn.Module.dump_patches = True and use the patch tool to revert the changes.
source code of class ‘torchvision.models.densenet._DenseLayer’ has changed. you can retrieve the

SourceChangeWarning: original source code by accessing the object’s source attribute or set torch.nn.Module.dump_patches = True and use the patch tool to revert the changes.

SourceChangeWarning: source code of class ‘fastai.layers.AdaptiveConcatPool2d’ has changed. you can retrieve the original source code by accessing the object’s source attribute or set torch.nn.Module.dump_patches = True and use the patch tool to revert the changes.

SourceChangeWarning: source code of class ‘fastai.layers.Flatten’ has changed. you can retrieve the original source code by accessing the object’s source attribute or set torch.nn.Module.dump_patches = True and use the patch tool to revert the changes.

Anyone knows how to solve this problem?

2

is that happening with a Learner that you exported with learn.export() ? If yes, simply reexporting should fix it.

3

Yes it is happening whit an exported learner. I will try the suggested fix. Thank you very much!