From the course: Full-Stack Deep Learning with Python
Unlock this course with a free trial
Join today to access over 23,400 courses taught by industry experts.
Preparing data for image classification using CNN - Python Tutorial
From the course: Full-Stack Deep Learning with Python
Preparing data for image classification using CNN
- [Instructor] Here I am on a brand new colab notebook, EMNIST classification using convolutional neural networks. Now, a new notebook implies a new colab runtime, which means we need to restart mlflow on this local machine. You can see from the message at the bottom that I've set up this runtime to run on a GPU. That's what we'll use to train our image classification convolutional neural network on the EMNIST data. Now, we need to install the libraries once again because we are on a new runtime. Torch, matplotlib, numpy, and pandas, we need all of these. We also need pytorch lightning because we'll be setting up our model using pytorch lightning, and we also need mlflow. Once again, this will get us mlflow 2.9.1, the latest version at the time of this recording. In order to ensure there are no breaking changes that mess up your demo, you might want to specifically install mlflow 2.9.1 yourself. And we also need pyngrok so…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.
Contents
-
-
-
-
-
(Locked)
Preparing data for image classification using CNN4m 2s
-
(Locked)
Configuring and training the model using MLflow runs6m 19s
-
(Locked)
Visualizing charts, metrics, and parameters on MLflow6m 52s
-
(Locked)
Setting up the objective function for hyperparameter tuning5m 35s
-
(Locked)
Hyperparameter optimization with Hyperopt and MLflow6m 21s
-
(Locked)
Identifying the best model3m 39s
-
(Locked)
Registering a model with the MLflow registry3m 12s
-
(Locked)
-
-