Advanced Usage
- Use PyNeuraLogic in combination with the PyTorch 🔥 framework.
- Learn how to represent heterogeneous graphs and possible ways to incorporate rules utilizing them into your templates.
- Evaluate your models directly in the database. Load data from a database, train a model, and transpile the model into SQL.
- PyNeuraLogic offers to utilize its engine only for inference as well. This section goes through an example to showcase the usage of the inference engine, to get all possible substitutions satisfying our queries.
- You can also extend the inference engine from the previous section and utilize numeric relations’ values. For example, to compute the shortest paths between points as in this example!
- Some relations can have special meanings and functionalities. You can find out more about them here.
- Having a visual representation of your model can help you get a better insight. Learn how to utilize prepared tools to visualize your models/templates and samples.
- Learn how recursive templates can be defined and utilized!
- In this section, we go through all the different settings of the backend engine, such as using its logging, debugging, passing additional JVM arguments, etc.