In this video, I will show you how you can train multiple neural networks on TPUs simultaneously. You can use this trick to train multiple folds for a dataset really quick and avoid all the optimization of hyperparameters that are usually associated with TPUs. I am not talking about how TPUs work.
Please note: you need to use "xm.optimizer_step(optimizer, barrier=True)" in the train_fn. This is not mentioned in the video.
You can see the full code here: https://www.kaggle.com/abhishek/super-duper-fast-pytorch-tpu-kernel
If you want to present something on my live show, fill up the form here: http://bit.ly/AbhishekTalks
#TPU #Tricks #DataScience
Follow me on:
Twitter: https://twitter.com/abhi1thakur
LinkedIn: https://www.linkedin.com/in/abhi1thakur/
Kaggle: https://kaggle.com/abhishek
Choose to Build with AI
Matched to Train a Team
AI Maker Residence 3
The third AI workshop taught by our legendary teacher, Nick Sarafa. This is a full-day hands-on training workshop for purposeful co-creation with AI using Claude Code.
Imagine having access to hundreds of billions of dollars of computing power and knowing exactly how to make it work for you through the power of super intelligence.
◆ Fri 09 Oct 2026◆ KOKO Cafe, London◆ With Nick Sarafa
