Tips N Tricks #6: How to train multiple deep neural networks on TPUs simultaneously
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
Видео Tips N Tricks #6: How to train multiple deep neural networks on TPUs simultaneously канала Abhishek Thakur
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
Видео Tips N Tricks #6: How to train multiple deep neural networks on TPUs simultaneously канала Abhishek Thakur
Показать
Комментарии отсутствуют
Информация о видео
Другие видео канала
![Building a summarizer using XGen-7b: Fully open source LLM by Salesforce](https://i.ytimg.com/vi/Y_93OghB8bY/default.jpg)
![TS-5: Automatic for the people](https://i.ytimg.com/vi/wBP8Pc4Wxzs/default.jpg)
![Talks S2E6 (Louise Ferbach): Deep Learning For Survival Analysis](https://i.ytimg.com/vi/-vmV4qA0ztQ/default.jpg)
![TS-4: Sales and demand forecasting](https://i.ytimg.com/vi/RdH8zd07u2E/default.jpg)
![TS-9: Hybrid methods for time series](https://i.ytimg.com/vi/NYZzBvKcfp4/default.jpg)
![Kaggle's 30 Days Of ML (Day-12 Part-2): Handling Categorical Variables](https://i.ytimg.com/vi/tRAczj-iQII/default.jpg)
![Tips N Tricks # 8: Using automatic mixed precision training with PyTorch 1.6](https://i.ytimg.com/vi/X7iOkhGePXg/default.jpg)
![Data representations for neural networks](https://i.ytimg.com/vi/-jCA0ctrLh0/default.jpg)
![Talks # 3: Lorenzo Ampil - Introduction to T5 for Sentiment Span Extraction](https://i.ytimg.com/vi/4LYw_UIdd4A/default.jpg)
![How to create GPT-powered conversational bot for any website](https://i.ytimg.com/vi/T1hdz3eU3bg/default.jpg)
![10K Subscribers: Approaching (almost) Any Machine Learning Problem and Talk Show](https://i.ytimg.com/vi/eXLcpsrcyGk/default.jpg)
![Bengali.AI: Handwritten Grapheme Classification Using PyTorch (Part-1)](https://i.ytimg.com/vi/8J5Q4mEzRtY/default.jpg)
![Kaggle's 30 Days Of ML (Competition Part-4): Hyperparameter tuning using Optuna](https://i.ytimg.com/vi/m5YSKPMjkrk/default.jpg)
![Kaggle's 30 Days Of ML (Day-11): Machine Learning Model to Predict House Prices (Intro to ML Ends)](https://i.ytimg.com/vi/IsVmI7HZ9DU/default.jpg)
![Panels # 1: What it takes to become a Quadruple Kaggle Grandmaster?](https://i.ytimg.com/vi/d7YvGX3oi3A/default.jpg)
![Best computer vision competitions on Kaggle (for beginners)](https://i.ytimg.com/vi/1-myowrUhok/default.jpg)
![Talks # 4: Sebastien Fischman - Pytorch-TabNet: Beating XGBoost on Tabular Data Using Deep Learning](https://i.ytimg.com/vi/ysBaZO8YmX8/default.jpg)
![Kaggle's 30 Days Of ML (Day-13 Part-2): Cross-validation](https://i.ytimg.com/vi/1zoxu3N94qU/default.jpg)
![Making Transformers go brum, brum, brum 🏎 (with Lewis Tunstall)](https://i.ytimg.com/vi/5E0nlHWgMMU/default.jpg)
![Multi-Lingual Toxic Comment Classification using BERT and TPUs with PyTorch](https://i.ytimg.com/vi/vvr_f-X_LaI/default.jpg)