TensorFlow: Save and Restore Models
Training a deep neural network model could take quite some time, depending on the complexity of your model, the amount of data you have, the hardware you’re running your models on, etc. On most of the occasions you’ll need to save your progress to a file, so in case of interruption (or a bug), you’ll be able to continue where you left off. Even more, after a successful training you’ll surely need to re-use the model’s learned parameters to make […]
Read more