1. What TensorBoard Is TensorBoard is mainly used to monitor how a model’s various metrics change — accuracy, loss, the weight distributions of each layer, and so on. TensorBoard is a visualization tool from TensorFlow that supports visualizing scalar, text, image, audio, video, and embedding data, among other types. But PyTorch can use TensorBoard too.

1. Create the Training Script Create a training script mnist.py with the following contents: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36