Package Base Details: caffe-git

Git Clone URL: https://aur.archlinux.org/caffe-git.git (read-only, click to copy)
Keywords: ai artificial cuda intelligence nvidia
Submitter: jyantis
Maintainer: dbermond
Last Packager: dbermond
Votes: 23
Popularity: 0.000000
First Submitted: 2014-12-08 01:10 (UTC)
Last Updated: 2022-03-01 14:46 (UTC)

Latest Comments

« First ‹ Previous 1 2 3 4 5 6 Next › Last »

mjal commented on 2016-01-20 14:37 (UTC)

For NVCC and boost1.6, another way is to add NVCCFLAGS=-D__STRICT_ANSI__ to Makefile.config (as seen on the arrayfire project https://github.com/arrayfire/arrayfire/issues/1210).

mjal commented on 2016-01-18 19:10 (UTC) (edited on 2016-01-20 14:40 (UTC) by mjal)

I also an issue with boost1.6 and the python binding (https://github.com/BVLC/caffe/issues/3494). EDIT: the linked pull request did the trick.

lostkhaos commented on 2016-01-11 18:49 (UTC) (edited on 2016-01-12 20:02 (UTC) by lostkhaos)

@chaitan94 I do believe that this is indeed an upstream problem with Boost. The problem is caused by NVCC attempting to use a float128, which is not supported right now (see: https://svn.boost.org/trac/boost/ticket/11852). For now, a safer patch than what you did: https://gist.github.com/anonymous/e680bfdb369baa52f08f I don't want to include this patch in this package, since it would modify a system-wide library.

chaitan94 commented on 2016-01-09 05:08 (UTC) (edited on 2016-01-09 05:27 (UTC) by chaitan94)

I get the following error on building as of 9th January 2015: /usr/include/boost/config/suffix.hpp(510): error: identifier "__float128" is undefined 1 error detected in the compilation of "/tmp/tmpxft_00007fe3_00000000-16_math_functions.compute_50.cpp1.ii". Makefile:565: recipe for target '.build_release/cuda/src/caffe/util/math_functions.o' failed make: *** [.build_release/cuda/src/caffe/util/math_functions.o] Error 1 ==> ERROR: A failure occurred in build(). Aborting... ==> ERROR: Makepkg was unable to build caffe-git. EDIT: I fixed the problem by changing the line __extension__ typedef __float128 float128_type; to __extension__ typedef long double float128_type; of the file /usr/include/boost/config/suffix.hpp

lostkhaos commented on 2015-12-01 18:57 (UTC)

This is a *-git package, please do not flag it as out-of-date.

jyantis commented on 2015-09-16 03:41 (UTC)

This has been fixed.

rtkt commented on 2015-09-01 21:08 (UTC)

python2-leveldb-svn package doesn't exist. Its PKGBUILD can be acquired from AUR3 archive https://github.com/aur-archive/python2-leveldb-svn/blob/master/PKGBUILD

davini commented on 2015-08-23 22:01 (UTC)

I get the following error building as of 8/23/2015 ... CXX/LD -o .build_release/tools/net_speed_benchmark.bin CXX tools/extract_features.cpp CXX/LD -o .build_release/tools/extract_features.bin .build_release/tools/extract_features.o: In function `int feature_extraction_pipeline<float>(int, char**)': extract_features.cpp:(.text._Z27feature_extraction_pipelineIfEiiPPc[_Z27feature_extraction_pipelineIfEiiPPc]+0x10f): undefined reference to `caffe::Net<float>::Net(std::string const&, caffe::Phase, caffe::Net<float> const*)' collect2: error: ld returned 1 exit status Makefile:562: recipe for target '.build_release/tools/extract_features.bin' failed make: *** [.build_release/tools/extract_features.bin] Error 1 ==> ERROR: A failure occurred in build(). Aborting... I haven't seen this error before, and I've built this package many times. Does anyone know what I'm doing wrong?

jyantis commented on 2015-08-21 18:33 (UTC)

gunterb14 - Fixed.

« First ‹ Previous 1 2 3 4 5 6 Next › Last »