You may think PyTorch performance tuning is a complex and daunting topic. This eBook breaks it down into easily consumable tips and tricks with concrete examples.
Discover the tuning tips that deliver optimal training speeds at lower costs. Reduce end-to-end latency by 5-10x, improve the accuracy of your model, and boost GPU utilization up to 90%.
In this book, you will dive deep into the training aspect of the machine learning pipeline. You will learn a set of optimizations and best practices that can accelerate model training in PyTorch. Presented techniques can be implemented by changing only a few lines of code and can be applied to a wide range of deep learning models across all domains.
In this comprehensive guide, you’ll learn:
- How PyTorch runs under the hood
- What can impact the performance of model training in the ML pipeline
- The process of optimizing PyTorch model training step-by-step
- 13 tuning tips including data loading, data operations, GPU processing, and CPU processing, with lines of code
- Real-world use cases using Alluxio as the data access layer for speed and efficiency
If you are managing analytics/SQL workloads, check out the optimization handbooks of Trino and Presto in the eBook series.
Complete the form below to access the ebook
.png)
Ebooks
Did you know major cloud service providers encourage you to put data in the cloud and charge you to get it out? Discover everything you need to know about data egress costs and never be surprised by a bill again.
An IDC survey¹ revealed that almost everyone (99%) incur planned or unplanned egress fees, and 41% incur these fees frequently. The impact of these charges is further highlighted by a study conducted by S&P Global², which found that 34% of enterprises have had to repatriate data on-premises or switch to a provider that doesn’t charge for egress.
Get the free ebook to learn:
- What egress fee is and why it matters
- AWS, Google Cloud and Microsoft Azure pricing model and the egress fees
- Best practices to help reduce egress costs up to 80%, including:
- Leveraging data caching to avoid roundtrips
- Streamlining data pipelines to minimize data replication
- Optimizing the data flow of your architecture
¹IDC, Future-proofing Storage: Modernizing Infrastructure for Data Growth Across Hybrid, Edge, and Cloud Ecosystem, March 2021.
²S&P Global Market Intelligence, Data center interconnection faces cloud-native competition, March 2022.
Your 🐰 queries are slow 🐢 … you’re frustrated 😩 …
Don’t let suboptimal Trino performance hold you back any longer!
Unlock the full potential of Trino and transform your data analytics game. Discover the secrets behind Trino’s query engine and learn how to overcome bottlenecks to achieve⚡ blazing-fast query performance.
In this comprehensive guide, you’ll learn:
- How Trino runs queries under the hood
- What bottlenecks can impact query performance in the query lifecycle
- How to refine Trino for optimal query performance
- Seven best practices for maximizing Trino query efficiency, including configuration settings, session properties, and SQL statements
- Real-world examples of Trino optimizations using caching
If you’re using Presto (PrestoDB), check out The Presto Optimization Handbook here.
If you are using Presto (PrestoDB) as the distributed query engine for data analytics, this comprehensive guide is your go-to resource for maximizing Presto’s potential on your data platform.
Get the best practices that have helped industry giants like Meta, Uber, and Walmart improve query performance by 3~10x. You will learn:
- How Presto query engine runs queries under the hood
- Identifying bottlenecks that impact query performance in the query lifecycle
- Refining Presto for optimal query performance
- Seven best practices for maximizing Presto query efficiency, including configuration settings, session properties, and SQL statements
- Presto optimizations at Uber scale and Fortune 1 scale
If you’re using Trino (formerly PrestoSQL), check out The Trino Optimization Handbook here.