You're reading from Practical Deep Learning at Scale with MLflow Bridge the gap between offline experimentation and online production

Product type Paperback

Published in Jul 2022

Publisher Packt

ISBN-13 9781803241333

Length 288 pages

Edition 1st Edition

Tools

MLflow

Concepts

Deep Learning

Author (1):

Yong Liu

Preface

1. Section 1 - Deep Learning Challenges and MLflow Prime

2. Chapter 1: Deep Learning Life Cycle and MLOps Challenges FREE CHAPTER

3. Chapter 2: Getting Started with MLflow for Deep Learning

4. Section 2 – Tracking a Deep Learning Pipeline at Scale

5. Chapter 3: Tracking Models, Parameters, and Metrics

6. Chapter 4: Tracking Code and Data Versioning

7. Section 3 – Running Deep Learning Pipelines at Scale

8. Chapter 5: Running DL Pipelines in Different Environments

9. Chapter 6: Running Hyperparameter Tuning at Scale

10. Section 4 – Deploying a Deep Learning Pipeline at Scale

11. Chapter 7: Multi-Step Deep Learning Inference Pipeline

12. Chapter 8: Deploying a DL Inference Pipeline at Scale

13. Section 5 – Deep Learning Model Explainability at Scale

14. Chapter 9: Fundamentals of Deep Learning Explainability

15. Chapter 10: Implementing DL Explainability with MLflow

16. Other Books You May Enjoy

Implementing a custom MLflow Python model

Let's first describe the steps to implement a custom MLflow Python model without any extra preprocessing and postprocessing logic:

First, make sure we have a trained DL model that's ready to be used for inference purposes. For the sake of learning in this chapter, we include the training pipeline MLproject in this chapter, so that we can easily produce a fine-tuned DL model. To run the training pipeline, make sure you have the virtual environment set up for this chapter by following the README file in this chapter's GitHub repository and set up the environment variables accordingly (https://github.com/PacktPublishing/Practical-Deep-Learning-at-Scale-with-MLFlow/blob/main/chapter07/README.md). Then, in the command line, run the following command to generate a fine-tuned model in the local MLflow tracking server:
```
mlflow run . --experiment-name dl_model_chapter07 -P pipeline_steps=download_data,fine_tuning_model
```

Once...

The rest of the chapter is locked