TypeError: ‘Adam’ object is not callable

#1

the file of this function is imported and when running (python train.py --epochs 1) i got this error:

--------------Training is starting------------- 
Traceback (most recent call last):
  File "train.py", line 48, in <module>
    suppFunctions.train_network(model, criterion, optimizer, trainloader, epochs, 20, power)
  File "/home/workspace/ImageClassifier/suppFunctions.py", line 128, in train_network
    loss = criterion(outputs, labels)
TypeError: 'Adam' object is not callable

the function train_network:

def train_network(model, criterion, optimizer, loader, epochs = 3, print_every=20, power='gpu'):
    
    steps = 0
    running_loss = 0

    print("--------------Training is starting------------- ")
    for e in range(epochs):
        running_loss = 0
        for ii, (inputs, labels) in enumerate(loader):
            steps += 1
            if torch.cuda.is_available() and power =='gpu':
                inputs, labels = inputs.to('cuda'), labels.to('cuda')

            optimizer.zero_grad()

            # Forward and backward passes
            outputs = model.forward(inputs)
            loss = criterion(outputs, labels)
            loss.backward()
            optimizer.step()

            running_loss += loss.item()

            if steps % print_every == 0:
                model.eval()
                vlost = 0
                accuracy=0


                for ii, (inputs2,labels2) in enumerate(vloader):
                    optimizer.zero_grad()
                    if torch.cuda.is_available():
                        inputs2, labels2 = inputs2.to('cuda:0') , labels2.to('cuda:0')
                        model.to('cuda:0')

                    with torch.no_grad():
                        outputs = model.forward(inputs2)
                        vlost = criterion(outputs,labels2)
                        ps = torch.exp(outputs).data
                        equality = (labels2.data == ps.max(1)[1])
                        accuracy += equality.type_as(torch.FloatTensor()).mean()

                vlost = vlost / len(vloader)
                accuracy = accuracy /len(vloader)



                print("Epoch: {}/{}... ".format(e+1, epochs),
                      "Loss: {:.4f}".format(running_loss/print_every),
                      "Validation Lost {:.4f}".format(vlost),
                       "Accuracy: {:.4f}".format(accuracy))


                running_loss = 0


    print("-------------- Finished training -----------------------")
    print("Dear User I the ulitmate NN machine trained your model. It required")
    print("----------Epochs: {}------------------------------------".format(epochs))
    print("----------Steps: {}-----------------------------".format(steps))
    print("That's a lot of steps")
#2

Hi,

Given the names, I guess you mixup the optimizer and the loss function? Here, criterion seem to contain your optimizer.

#3

@albanD thank you for your replay, i think i figured out where is the problem! but i don’t know how to solve it…

i have uploaded the files to make it easy to refer :
link

error said that ‘Adam’ object is not callable and the error in is in line 128 in train_network from suppFunctions.py file

in the train.py line 45 we create an optimizer inctance from nn_setup fucntion in suppFunctions.py
i saw that in line 103 from the suppFunctions.py this line of code:

optimizer = optim.Adam(model.classifier.parameters(), lr )

and when i read the doc of pytorch i figured that i passed a wrong parameters
could you help me writing the file in right way ?

#4

The problem is that here you return model, criterion, optimizer
But here you unpack model, optimizer, criterion.

Notice that the optimizer and criterion are swapped. You need to swap them back.

1 Like
#5

@albanD Thank you alot it worked even if i miss types some variable names, but now can i ask you how to run predict.py file ?
i run train.py by typing python train.py --epochs 1 in my work-space command line

#6

It looks like a standalone file no?
I don’t know.

1 Like