From the course: Building Deep Learning Applications with Keras
Join today to access over 24,800 courses taught by industry experts.
Visualizing computation graphs with TensorBoard
From the course: Building Deep Learning Applications with Keras
Visualizing computation graphs with TensorBoard
- [Instructor] We are going to take a tour through a neural network's training process, continuing to use the TensorBoard. This time, we are going to select the graphs tab. Graphs are our guide, understanding the intricate workings of a neural network. Much like a city map, it's like navigating through the city map's streets. What lies before us is nothing short of a map of our model's neural pathways we just created, the computational graph that Keras has constructed. Now, each node is an operation or a variable, and the lines represent the flow of the data. Much like roads connecting different locations in a city, the graph maps out the journey our data takes from input to the output. Now, let's find the main avenue in the city map. Which one do you think that is? It's nothing but the sequential. The sequential part of the graph is like the main avenue of our neural network city. It's a series of stops or layers that we have named, remember, through which our data travels. Each…
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.