Using model.pth pytorch to predict image

in my database, there are 26 types of objects.

is there a possibility I did one test by giving 1 type of image then compared to all of my data and outputting the closest range ??

It’s ok, but you need another one in the output

I do not understand.

can you explain it

or can you give an example?

pleaseeee

Just use prob = nn.Sigmoid()(output) and you’ll get the similarity between the inputs as a probability

oke.
Thankyou so much,
:laughing: :laughing: :laughing: :laughing: :laughing:

Hi, you could use this code as a blueprint as well.

Thanks

Michael