Age | Commit message (Collapse) | Author | Files | Lines | |
---|---|---|---|---|---|
2014-06-26 | Merge pull request #488 from longjon/wall-werror | Evan Shelhamer | 1 | -3/+4 | |
Build with -Wall -Werror | |||||
2014-06-25 | Merge pull request #478 from kloudkl/cpu_only_tests | Evan Shelhamer | 1 | -1/+6 | |
CPU only tests | |||||
2014-06-25 | move CUDA 6.0 check into switch statement itself | Jonathan L Long | 1 | -0/+2 | |
This allows -Wswitch to be turned on so that the compiler can check exhaustiveness. | |||||
2014-06-25 | initialize in declared order | Jonathan L Long | 1 | -3/+2 | |
2014-06-20 | Rename curand_availability_logged according to the Google style guide | Kai Li | 1 | -3/+3 | |
2014-06-20 | Suppress redundant log messages of unavailable curand | Kai Li | 1 | -1/+6 | |
2014-06-20 | Initialize the RNG generator with an orthogonally newed Generator | Kai Li | 1 | -1/+1 | |
2014-06-16 | in Caffe::SetDevice, call cudaSetDevice before Get | Jonathan L Long | 1 | -1/+3 | |
Otherwise initialization will be performed on whichever device is default. | |||||
2014-05-13 | add cublas status in cuda 6 to fix warning | Evan Shelhamer | 1 | -0/+2 | |
...and #define for older CUDAs to not break the build. | |||||
2014-04-28 | rollback 8368818, does not build | Jeff Donahue | 1 | -2/+0 | |
2014-04-28 | Handling CUBLAS_STATUS_NOT_SUPPORTED to suppress warning | Sergey Karayev | 1 | -0/+2 | |
2014-04-18 | remove now unused set_generator and related code | Jeff Donahue | 1 | -14/+0 | |
2014-04-18 | pass caffe rng ref into variate_generator constructor instead of having | Jeff Donahue | 1 | -3/+3 | |
caffe rng adopt its state | |||||
2014-04-18 | remove unnecessary return from void set_generator | Jeff Donahue | 1 | -1/+1 | |
2014-04-08 | make rng_ a private member of Generator | Jeff Donahue | 1 | -1/+3 | |
2014-04-08 | call caffe_set_rng at the end of each vRng function to maintain state | Jeff Donahue | 1 | -20/+20 | |
2014-04-02 | Merge pull request #278 from kloudkl/log_error_string_in_check_macro | Jeff Donahue | 1 | -0/+54 | |
Log error strings in the error/status checking macros | |||||
2014-04-01 | seed boost rng with cluster_seedgen by default | Jeff Donahue | 1 | -1/+3 | |
2014-04-01 | Fix cpplint errors | Kai Li | 1 | -1/+1 | |
2014-04-01 | Add curandGetErrorString and use it to redefine CURAND_CHECK | Kai Li | 1 | -0/+32 | |
2014-04-01 | Add caffe::cublasGetErrorString and redefine CUBLAS_CHECK with it | Kai Li | 1 | -0/+22 | |
2014-03-22 | Set copyright to BVLC and contributors. | Evan Shelhamer | 1 | -2/+1 | |
The exact details of the contributions are recorded by versioning. | |||||
2014-03-22 | Hide boost rng behind facade for osx compatibility | Evan Shelhamer | 1 | -1/+37 | |
Split boost random number generation from the common Caffe singleton and add a helper function for rng. This resolves a build conflict in OSX between boost rng and nvcc compilation of cuda code. Refer to #165 for a full discussion. Thanks to @satol for suggesting a random number generation facade rather than a total split of cpp and cu code, which is far more involved. | |||||
2014-03-21 | lint | Evan Shelhamer | 1 | -2/+1 | |
2014-03-21 | clean up residual mkl comments and code | Evan Shelhamer | 1 | -13/+1 | |
The FIXMEs about RNG were addressed by caffe_nextafter for uniform distributions and the normal distribution concern is surely a typo in the boost documentation, since the normal pdf is correctly stated elsewhere in the documentation. | |||||
2014-03-21 | compile caffe without MKL (dependency replaced by boost::random, Eigen3) | Rodrigo Benenson | 1 | -9/+14 | |
- examples, test and pycaffe compile without problem (matcaffe not tested) - tests show some errors (on cpu gradient tests), to be investigated - random generators need to be double checked - mkl commented code needs to be removed | |||||
2014-02-26 | long -> int64_t; all linter errors fixed. woohoo! | Jeff Donahue | 1 | -2/+2 | |
2014-02-26 | fix most linter errors | Jeff Donahue | 1 | -2/+3 | |
2014-02-03 | Added device_query.cpp to examples/ to get basic information about the ↵ | sguada | 1 | -0/+1 | |
current GPU device or other device_id | |||||
2013-11-16 | removed the StillFresh function that was used to make sure training does not ↵ | Yangqing Jia | 1 | -16/+0 | |
happen with an earlier version of code. | |||||
2013-11-07 | distributed solver still having bugs. Pausing for now... | Yangqing Jia | 1 | -3/+13 | |
2013-10-31 | solver restructuring: now all prototxt are specified in the solver protocol ↵ | Yangqing Jia | 1 | -0/+18 | |
buffer | |||||
2013-10-30 | common.cpp: rand seed fix | Yangqing Jia | 1 | -2/+2 | |
2013-10-28 | caffe common cpp: fixed an embarassing bug | Yangqing Jia | 1 | -4/+7 | |
2013-10-24 | pushing missing checkout | Yangqing Jia | 1 | -2/+2 | |
2013-10-24 | common.cpp rand seed: do time(NULL) so we can have different initializations | Yangqing Jia | 1 | -2/+2 | |
2013-10-24 | common: added DeviceQuery() function | Yangqing Jia | 1 | -0/+30 | |
2013-10-17 | freshness | Yangqing Jia | 1 | -0/+18 | |
2013-10-10 | cpplint | Yangqing Jia | 1 | -2/+6 | |
2013-10-07 | common.hpp/cpp update, and lenet using random subcrop | Yangqing Jia | 1 | -48/+24 | |
2013-09-27 | cpplint | Yangqing Jia | 1 | -1/+1 | |
2013-09-23 | copyright message | Yangqing Jia | 1 | -0/+2 | |
2013-09-22 | naming. I might regret it someday. | Yangqing Jia | 1 | -0/+76 | |