My Journey to PyTorch
summary
Starts as a user of NumPy-like PyTorch in 2017, moves from debugging a custom segmentation loss to landing a merged improvement to torch.split feature, then joins NVIDIA to support GPU users. Reports specific community impact: 20,000 forum posts and a forum with 3 million visitors per month. Notes practical ecosystem progress: domain libraries (HuggingFace, PyTorch Geometric), TorchScript and mobile backends, and uses forum feedback to prioritize performance and usability tradeoffs.
AI-written summary · about
chapters (5)
- 0:00Introduction
- 0:25Ecosystem Day overview
- 1:43Piotr's journey to PyTorch
- 5:55PyTorch growth and features
- 7:25Future and closing
related talks
12mKeynote: How PyTorch Became the Foundation of the AI Revolution
47mGet to Know PyTorch - A Cutting Edge Open Source AI Framework for Deep Learning
12mLightning Talk: State of PyTorch
48mFireside Chat
43mPyTorch Technical Deep Dive
54mKeynote: PyTorch 2.1 Technical Deep Dive
32mKeynote: Fireside Chat with Soumith Chintala, Co-Creator of PyTorch & Merve Noyan
19mGet to know PyTorch - a cutting edge open source AI framework for Machine Learning
