Deepchecks has collaborated with W&B so that you can test and validate your ML models as part of the experiment tracking workflow. This is extremely exciting since we believe that both experiment tracking and testing are “best practices” that every ML practitioner should be using, so combining these tools lowers the barrier to building models as we all should.
Want to see how it works? Watch the webinar to learn about:
- Types of problems that ML models face (i.e. data integrity and distributions, ML methodology pitfalls, model evaluation related issues)
- Best practices for validation: when, how and what to test for
In addition, we’ll present a live demo of using the deepchecks open-source package for effective validation with minimal effort, and present the recent option to integrate the validation’s results with W&B.