i combined cnn(squeezenet)+lstm for liver classification i got 91% as train accuracy and 0.276 for loss…but i want to know why the result of my conf max is 0 everywhere, please help!
classes = [‘Hcc’, ‘Meta’, ‘Hema’, ‘Heal’]
out_dir = “./”
set_name = "Liver Lesion’’
model = combine()
model. load_state_dict(torch.load(’/gk4_net.pth’))
y_pred, y_true, y_prob = get_output(model, testloader)
cnf_matrix = confusion_matrix(y_pred, y_true, labels=np.arange(4))
show_confMat(cnf_matrix, classes, set_name, out_dir)