This section explains the process used to access the MNIST datasets:
- Once we receive a confirmation that the TensorFlow library has been properly installed, it is imported into the notebook.
- We can confirm the version of TensorFlow as well as extract the images to our local folder of MNIST/. The extraction process is visible in the output of the notebook, as seen in the following screenshot:
- The four extracted files are named the following:
- t10k-images-idx3-ubyte.gz
- t10k-labels-idx1-ubyte.gz
- train-images-idx3-ubyte.gz
- train-labels-idx1-ubyte.gz
- They have been downloaded to the MNIST/ subfolder as seen in the following ...