From the course: Full-Stack Deep Learning with Python

Unlock this course with a free trial

Join today to access over 24,700 courses taught by industry experts.

Configuring and training the model using MLflow runs

Configuring and training the model using MLflow runs - Python Tutorial

From the course: Full-Stack Deep Learning with Python

Start my 1-month free trial Buy for my team

Configuring and training the model using MLflow runs

“

- [Instructor] We can now set up our convolutional neural network for image classification. Once again, I import some PyTorch libraries and classes that we'll need to set up our network. And next we'll create a new experiment within which we'll have a training run for our cnn. I call mlflow.create.experiment. The name of the experiment is emnist_letters_prediction_using_cnn, and I also set this as the active experiment so that any run that we create will be within this experiment. Next, we set up a class that inherits from a lightning module where we'll define our convolutional neural network, as well as the training and validation steps for our model. As before, for our classification model, we'll use the cross entropy loss function. Next, we define our convolutional neural network with multiple blocks of convolutional and polling layers. On line 8, we have our first convolutional layer. It takes in a single channel…

Contents