Learning Deep Learning with MinPy¶
We believe there is no easy shortcut if one wishes to master the concepts of deep learning and make novel contributions. A very nice introduction to deep learning is CS231n from Stanford, which contains self-paced lectures and notes. This is a great course, particularly because of its hands-on homeworks.
We tailor-made a modified version of the coursework using MinPy, for assignment 2 and 3, fixing exactly the two problems that the original NumPy coureware lacks: the support to run on GPU and auto-gradient. The new homework can be found here: CS231n Homework links
This new courseware has been pilot in ShanghaiTech and Shanghai JiaoTong universities. We want to thank Karpathy for giving the permission, and the instructors for trying it out. Please send us feedbacks if you run into issues.