PyTorch Upgrades to Cloud TPUs, Links to R
A version of the PyTorch machine learning framework that incorporates a deep learning compiler to connect the Python package to cloud Tensor processors (TPUs) is now available on Google Cloud, the public cloud vendor and PyTorch co-developer Facebook announced.
The general availability on PyTorch/XLA means users can access cloud TPU accelerators via a stable integration, the companies said Tuesday (Sept. 29).
Separately, promoters of the programming language R released a package that allows developers to use “PyTorch functionality natively from R.” The new tool, dubbed “Torch for R,” requires no Python installation.
Meanwhile, Facebook and Google said PyTorch/XLA combines the machine learning library’s APIs with XLA’s linear algebra compiler that targets CPUs, GPUS and, now, cloud TPUs. While running on most standard Python programs, PyTorch/XLA defaults to CPUs for operations not yet supported on Tensor processors.
That framework helps PyTorch users “find bottlenecks and adapt their programs to run more efficiently on cloud TPUs,” said Craig Wiley, director of product development for Google’s Cloud AI platform.
The company cited several PyTorch/XLA use cases that tap into cloud TPUs to accelerate projects. Those use cases range from training neural networks to upgrading language models with visual components.
Google Cloud also said it is supporting open source implementations of deep learning models to foster greater use of PyTorch/XLA, including ResNet-50 and the Deep Learning Recommendation Model. It is also developing open source tools for continuous testing of machine learning models on cloud TPUs.
Google Cloud released TPUs on its machine learning engine more than two years ago as part of its “AI first” strategy. The ML engine released in 2017 is a managed service for accessing the TensorFlow open source computational library.
Elsewhere in the PyTorch ecosystem, R Studio said this week its release of “R for Torch” responds to community advocacy for broader support of PyTorch. Along with a “leaner software stack” to reduce installation issues, R Studio said Torch for R does not require installation of a Python environment to achieve desired functionality via R.