Sequential Model - TensorFlow Essentials #1

Key Takeaways

hi welcome to this introduction to the sequential model in tensorflow so we're just going to go ahead and start a new notebook here now i want to do is import tensorflow now the sequential model is one of two different approaches that we can use for building our models in tensorflow those are the functional api and then the sequential model that we're going to go through here so the sequential model is better for simple sequential stacks of layers where each layer has just one input and just one output so if our model looks something like this then the sequential approach is probably better so we access the sequential model using tf.keras.sequential so it looks like this and then that initializes the sequential model and then there are two methods for adding layers with this model the first of those which i'm going to show you is using the add method so all we do is take model add and then we add our layer here so one of the layers that we'll be using quite a lot is a densely connected neural network layer so we just write tf keras layers dense and then we're going to add in 32 units here and our input shape will be 10 units so here we have 10 inputs and they are each connected to 32 units within our dense layer here and they will use a rayleigh activation function which is a rectified linear unit now you don't need to know all this right now but what i do want you to focus on is the shape that we are producing so if we go over to here this is the shape of the network that we're building so we have 10 inputs so that was represented by the input shape and then we have 32 neurons in the middle so 32 units if you look because it's a density connected neural network every single one of these is connected to every single one so all of our 10 inputs are connected to this unit here this one this one this one and so on and then in our output i want to put two units so we'll go ahead and do that now so if we want to add another layer we just call me add again so we can run that okay sorry this is input shape and if we want to add another layer all we do is model add again now we only need to define the input shape in our first layer with the sequential model otherwise we don't need to define it because it will assume that our input shape matches the shape of the previous layer which in this case is 32 so we don't need to do anything there so we just put two and then we add in our activation now because this is the output layer and we have more than one output we will be using the softmax activation so in the output layer you will usually have either sigmoid or softmax if you just have one output so if your output is just one value between zero and one you use sigmoid if you have multiple outputs so we have two units here that means our output can look something like this and what we will want to do is take the maximum value as the answer so in this case it would be indexed at zero so the output label would be zero and when we're doing that we want to use the softmax activation function so we can execute that and then we can print out our model with the model summary and here we can see okay we have a sequential model and it has two layers so this doesn't include the input layer so our input layer is actually 10 units and then it goes into the actual model which is 32 units this is our hidden layer and then our output layer is two units at the end there and we see that there are a total of 418 parameters and all those trainable so that those are the number of numbers that the model will adjust in order to learn patterns within the training data okay so that is our first option for building a sequential model but we also have another way of building it so rather than using the add method we can actually define our model all within the single initialization function like this we just add a list and then we take what we had here just the layer and add it in and you just add each layer as a item within the list and then we just execute that print the model summary and we have the exact same model so that is everything on the sequential model i hope you've enjoyed and i hope it's been useful so thank you for watching and i will see you again soon bye