Conclusion

Conclusion

In this lesson, you learned about two popular formats for machine learning: TensorFlow and PyTorch. You saw the most common use cases for these platforms and how their model files differ. You then set up and learned how to use conda to build isolated environments you can use for Python libraries. You installed Core ML Tools into the environment and then installed PyTorch to prepare the way for conversion of a PyTorch model to Core ML.

If you’ve gotten the impression that working with machine learning models can be complicated, you’re correct. The steps you’ve done in this first section will make your work easier, but you’ll find quirks and odd issues whenever you work with these models due to their complexity and interdependence upon so many external libraries. In the next lesson, you’ll learn about model conversion and convert your PyTorch model to CoreML before using it in an iOS app.