shurans
fc6.tensor无法生成，直接在终端执行，会提示核心已转储
I encountered same problem, when I run the dss_marvin_script(0,100,1,[] ,1,'RPN_NYU',1,[],0,0); I found we can not get the file 'fc6.tensor'
The author mentioned To run the code it requires around 12 G GPU memory, and there is no obvious way to fix it.
发件人: zhoufengbuaa [email protected] 发送时间: 2016年9月1日 15:09 收件人: shurans/DeepSlidingShape 抄送: Tarssy; Mention 主题: Re: [shurans/DeepSlidingShape] fc6.tensor无法生成，直接在终端执行，会提示核心已转储 (#12)
did you solve this problem yet? @Tarssyhttps://github.com/Tarssy
― You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/shurans/DeepSlidingShape/issues/12#issuecomment-243994715, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AT50OebHDMxG0YWRSTSyodbdWV5QY9eIks5qlnoYgaJpZM4JtDNu.
@Tarssy @zhoufengbuaa Hi, I've encountered the following problems but I don't know how to fix it. Also, what do you mean by "直接在终端运行" ? Will that fix the former issue?
> Error using feof
> Invalid file identifier. Use fopen to generate a valid file identifier.
>
> Error in readTensor (line 11)
> while ~feof(fp)
>
> Error in orgnizefeat (line 38)
> features_fc6 = readTensor([toreadfeaturefolder 'fc6.tensor'],true);
>
> Error in dss_marvin_script (line 262)
> orgnizefeat(featurefolder,featureLocalfolder,outfilename_p,cls,opt);
It will be very nice of you to tell me how to fix this problem. Thanks.
FYI, I've run the code on 12G GPU.
This problem exist because the marvin command is not successfully run though matlab. You can try to copy out the command start with "cd .... " to terminal to run marvin directly.
@Tarssy @TyaCUI I have the same error as yours. I run the code on 12.2G GPU and I used command line "cd..." and "...train.sh". Is the space enough or did I do something wrong? Thank you so much if you reply the question.
Only on Graphics card with around 12G memory can the code run successfully?
Nvidia gtx 1080ti is 11G memory. Can I run the code successfully on 1080ti? Thanks.
