I was trying to use fastai library for multi class image classification. The labels were initialized with a dataframe. I got the following output as result during testing.
Hi,
I think its most likely an error because of incorrect labeling.
Can you check how many total output classes does the model have? And does it match with the actual number of classes? Some classes must be repeating…
1 Like
Thanks for the reply. The model has 3 output classes. Actual model has 14 classes. I used its weights to load the learner.
Yeah, so your model thinks there are 14 classes. Thats why you’re getting the same output prediction thrice.
Can you help me know how to change it ? The below is my data loader code
