The tensor engine was first implemented inside SN3 (before it was called Lush) in 1992 at Bell Labs ...
TL;DR · AI 摘要
Yann LeCun回顾1992年在贝尔实验室与Léon Bottou开发的SN3（后称Lush）中首次实现张量引擎，其命名规范延续至PyTorch等现代深度学习库。
核心要点
- 张量引擎最早于1992年在贝尔实验室的SN3/Lush系统中实现
- PyTorch等库沿用了该早期系统的张量操作命名规范
- EBlearn、Torch5/7等项目继承并传播了这一技术传统
The naming convention has survived to this day in PyTorch and other libraries.
The naming of the tensor operations was reused in EBlearn (C++ deep learning" / X
Yann LeCun on X: "@norpadon The tensor engine was first implemented inside SN3 (before it was called Lush) in 1992 at Bell Labs by Léon Bottom and me. The naming convention has survived to this day in PyTorch and other libraries. The naming of the tensor operations was reused in EBlearn (C++ deep learning" / X
Don’t miss what’s happening
The tensor engine was first implemented inside SN3 (before it was called Lush) in 1992 at Bell Labs by Léon Bottom and me. The naming convention has survived to this day in PyTorch and other libraries. The naming of the tensor operations was reused in EBlearn (C++ deep learning library written by Pierre Sermanet and me, with some help from
). It was recycled in Torch5 and Torch7, which was written largely by Ronan Collobert, and my students Clément Farabet, and
). Clément and Koray had been brought up on Lush (the open version of SN) and knew the nomenclature. Then, Soumith used the same conventions in PyTorch.
·
15
48
659
189
Read 15 replies