From the course: Full-Stack Deep Learning with Python
Join today to access over 24,400 courses taught by industry experts.
Exploring parameters and metrics in MLflow - Python Tutorial
From the course: Full-Stack Deep Learning with Python
Exploring parameters and metrics in MLflow
- [Instructor] With model training complete, let's head over to the MLflow UI, and if you hit Refresher, you'll find that our new experiment containing our PyTorch model run, is available. Here we have the emnist_letters_prediction_using_DNN experiment. And within this experiment, we have a single run, that is the training of our PyTorch model that we just completed. Now let's take a look at the charts first, because they're very interesting. You'll see that MLflow automatically logs the charts, for all of the metrics that were tracked during training. Here you can see first a bar chart, tracking the accuracy on the test data. That's 0.85. If you scroll down, you'll find other useful charts. Let me show you a few. Here is a chart of the training loss over the epoch of training. You can see how the loss falls as we train for longer. Here is a chart for the validation accuracy, which grows as we train for longer,…
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.
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.
Contents
-
-
-
-
Loading and exploring the EMNIST dataset4m 53s
-
(Locked)
Logging metrics, parameters, and artifacts in MLflow6m 6s
-
(Locked)
Set up the dataset and data loader3m 47s
-
(Locked)
Configuring the image classification DNN model4m 56s
-
(Locked)
Training a model within an MLflow run4m 6s
-
(Locked)
Exploring parameters and metrics in MLflow4m 19s
-
(Locked)
Making predictions using MLflow artifacts5m 21s
-
-
-
-