My portfolio of PyTorch writings

My work as a developer advocate leads me to meet several interesting people and ideas, and in the PyTorch community there is no dearth of either. Here I share a sampling of the content I create to make using PyTorch easier, or amplify inspirational voices in the community.

PyTorch 101:

• Intro to PyTorch @ LFSummit
• Learn the basics
• Microsoft AI Show

Twitter Microblogs:

• Low Numerical Precision in PyTorch
• Tensor Memory Formats
• Inference with Quantization

Community Interviews:

• Refik Anadol creates mind-warping art with PyTorch
• Autodesk uses PyTorch to build a production-scale chatbot
• A fun PyTorch-based library to blend visual aesthetics
• Training and Inference at scale with Ray
• Qualcomm’s AIMET enables compressing PyTorch models for edge devices
• Quantum(!!) Machine Learning in PyTorch
• PyTorch-based TorchIO helps medical image preprocessing and augmentation
• Multi-task reinforcement learning library in PyTorch

full list

Performant PyTorch:

• Profiling PyTorch performance
• Leaner and Greener AI with Quantization in PyTorch
• A cookbook for Quantization in practice
• Efficient AI Training and Inference with PyTorch
• Summer of open source: building more efficient AI with PyTorch
• Efficient PyTorch: Tensor Memory Format Matters

Distributed Training Video Tutorials:

1. Part 1: Welcome to the Distributed Data Parallel (DDP) Tutorial Series
2. Part 2: What is Distributed Data Parallel (DDP)
3. Part 3: Multi-GPU training with DDP (code walkthrough)
4. Part 4: Multi-GPU DDP Training with Torchrun (code walkthrough)
5. Part 5: Multinode DDP Training with Torchrun (code walkthrough)
6. Part 6: Training a GPT-like model with DDP (code walkthrough)

Workshops:

• Quantization in PyTorch
• Deep Learning and Generative AI Workshop @ MIT Sloan (code)

Other blogs:

• Building a deep learning karaoke application with PyTorch
• Case Study of Amazon Ads using PyTorch in production