Keynote: The Promise of PyTorch as a General-Purpose Array-Oriented Computational..
summary
Argues that PyTorch now serves as a general-purpose array backend by implementing the Array API, so libraries like scikit-learn and SciPy can dispatch to a PyTorch backend with only small upstream changes and an Array API compatibility layer. Quansight and Meta integration lets NumPy-style code run on CPU, GPU, or TPU, and examples show existing scikit-learn pipelines switching to torch arrays and achieving much faster execution with minimal code changes.
AI-written summary · about
related talks
36mScikit-learn on GPUs with Array API
19mGet to know PyTorch - a cutting edge open source AI framework for Machine Learning
28mA Hitchhiker's Guide to the Array API Standard Ecosystem
32mKeynote: AI & the stuff built for AI - are they actually useful for data science?
26mFrom NumPy to PyTorch, A Story of API Compatibility
19mIncreasing the Impact of AI Through Better Software (Pytorch Updates)
29mKeynote
32mKeynote: Fireside Chat with Soumith Chintala, Co-Creator of PyTorch & Merve Noyan