Mastering TensorFlow 1.x

The tfruns package in R

You can follow along with the code in the Jupyter R notebook ch-17d_TensorBoard_in_R.

The tfruns package is very useful tool provided in R that helps in tracking multiple runs for training the models. The run data is automatically captured by the tfruns package for models built using the keras and tfestimators packages in R. Using tfruns is pretty simple and easy. Just create your code in an R file and then execute the file using the training_run() function. For example, if you have a mnist_model.R file, then execute it using the training_run() function in the interactive R console as follows:

library(tfruns)training_run('mnist_model.R')

Once the training is finished, the window displaying the summary of the run appears ...