Guest
Launch TensorBoard for Keras model directly from Jupyter Notebook and check the training performance (loss, accuracy graphs) in real time on your broswer. Also check the architecture of your defined neural network in user friendly way on Tensorboard schema very simply. For this example, I used Multivariate Time Series Prediction model from here written in Python by using LSTM architecture: Just add Tensorboard callback from Keras and some magical functions to your Jupyter Note
