Jupyter Notebook C++ Python Cuda Objective-C++ CMake Other
Latest commit 3ebce25 Apr 27, 2017 Kevin Matzen committed with facebook-github-bot fix curand odd-sized workaround
Summary: Ran into illegal memory access errors when running MSRAFill on an odd-sized tensor.  curand only supports even-sized fills.  To workaround this limitation, we fill the last entry of the tensor manually and use curand for what remains.  In this line, the intent is to get the n-1 th element of the tensor.  r is already a T*, so we should not be multiplying by sizeof(T) to get the n-1 th element.

Differential Revision: D4961306

fbshipit-source-id: 587f2945abf025e28f573482a4828c09e6ae771b



Caffe2 is a lightweight, modular, and scalable deep learning framework. Building on the original Caffe, Caffe2 is designed with expression, speed, and modularity in mind.

Questions and Feedback

Please use Github issues (https://github.com/caffe2/caffe2/issues) to ask questions, report bugs, and request new features.

Please participate in our survey (https://www.surveymonkey.com/r/caffe2). We will send you information about new releases and special developer events/webinars.

License and Citation

Caffe2 is released under the BSD 2-Clause license.

Build Status

Travis Build Status Windows Build status

Detailed build matrix (hit refresh if you see icons not showing up due to heroku):

Target Status
Linux Build Linux
Mac (CPU) Build Mac
Android Build Android
iOS Build iOS
Linux + MKL Build LinuxMKL
Windows Build status

Further Resources on Caffe2.ai