From the course: Applied AI: Getting Started with Hugging Face Transformers
Join today to access over 24,800 courses taught by industry experts.
Reviewing model configuration
From the course: Applied AI: Getting Started with Hugging Face Transformers
Reviewing model configuration
- [Instructor] In the previous video, we explored the model architecture. We will also explore additional metadata about the model now. By printing the model.config attribute, we get a dictionary of configuration information about the model. Let's explore what it shows for the NER model. The name shows the name of the model. The num label shows the number of labels or classes predicted, the architecture used is built for token classification. A number of configuration parameters represent hyper parameters used for training. Of specific interest is the ID to label parameter. It shows the nine different classes predicted. A zero is predicted if the given token does not represent any entity type. The rest of the classes are self-explanatory. Once again, exploring this configuration is particularly useful if we need to use transfer learning for customizing the model.
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.