Closed zhuotest closed 6 years ago
Thank you for submitting the issue! @sandeep-krishnamurthy requesting this be labeled as installation.
Finally I figured this out. Its my incorrect configuration. I should have installed newer GPU Card driver since I installed newer version of CUDA (9.2). In fact I installed nvidia-396 on my 1080Ti.
Hey, how did you get the nvidia-396 driver?
@zhuotest you can confirm But @rohun-tripathi does this help - https://www.nvidia.com/drivers/beta ?
Note: Providing complete information in the most concise form is the best way to get help. This issue template serves as the checklist for essential information to most of the technical issues and bug reports. For non-technical issues and feature requests, feel free to present the information in what you believe is the best form.
For Q & A and discussion, please start a discussion thread at https://discuss.mxnet.io
Description
I'm on ubuntu16.04 with GPU and pip, and follows instruction in official documentation web page (https://mxnet.incubator.apache.org/install/index.html?platform=Linux&language=Python&processor=GPU). I follow that instructions, which can install mxnet-cu92 but cannot run example code correctly.
Environment info (Required)
Package used Python2.7.12
Build info
I first download and intalled cuda-9.2 and cudnn7.1 from nvidia website. Then I
sudo pip install mxnet-cu92
Error Message:
see next section for full info
Minimum reproducible example
I use the official website's example code in a interpret envirionment. The code and output is:
Steps to reproduce
install ubuntu16.04 install cuda9.2 install cudnn7.1 use python 2.7.12 Open terminal and type:
What have you tried to solve it?
Note that I can build official Caffe with my installed cuda9.2 and cudnn7.1 (https://github.com/BVLC/caffe). It only gives some cudnn warnings but can complete compilation.