Dear Community I am working on landmark classification problem but during my training of using a convolutional neural network from scratch by using about 8 layers of CNN in the backbone and 4 layers of MLP in the head i.e 3 hidden layers and 1 output layer during my training my training loss drops till it get stuck at 3.9 my validation loss starts from 3.9 and get stucked there I am using sgd optimizer but it seems that i got trapped at a local minimum any ideas to improve this network
You posted this in DLS Course 1 (Neural Network and Deep Learning) but there is no such assignment that uses CNN. Please move your post to the correct course category as described here and our mentors will be happy to assist you.