TFLearn hides many of the implementation details exposed by TensorFlow and, in many cases, it allows us to focus on the definition of the ConvNet with a higher level of abstraction. Our pipeline reaches an accuracy of 88 percent in 50 iterations. The following image is a snapshot of the execution in a Jupyter notebook:
![](http://images-20200215.ebookreading.net/16/3/3/9781788293594/9781788293594__tensorflow-1x-deep__9781788293594__assets__0519e02a-cae4-4ef6-9fa5-b05a2eb880b7.png)
An example of Jupyter execution for CIFAR10 classification