Machine-Learning-Collection Image Captioning gives following error: TypeError: relu(): argument 'input' (position 1) must be Tensor, not InceptionOutputs

Image Captioning gives following error: TypeError: relu(): argument 'input' (position 1) must be Tensor, not InceptionOutputs

Open JoshFruin opened this issue 2 years ago • 1 comments

Aug 28 '23 14:08 JoshFruin

Inception net gives 2 outputs. The second output is auxiliary output. So you might want to try - features = self.inception(images)[0]

Apr 02 '24 17:04 Aditya2814