1. What is TensorBoard TensorBoard is mainly used to monitor the changes of various metrics of the model, such as accuracy, loss, weight distribution of various layers, etc. TensorBoard is a visualization tool for TensorFlow, but PyTorch can also use TensorBoard. TensorBoard is a visualization tool for TensorFlow that supports scalar, text, image, audio, video, and Embedding data visualizations, but PyTorch can use TensorBoard as well. 2. Install tensorboard 1 pip install tensorboard 3 . use