Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

nd convolution and pooling with cuDNN #3983

Open
wants to merge 32 commits into
base: master
Choose a base branch
from
Open

nd convolution and pooling with cuDNN #3983

wants to merge 32 commits into from

Commits on Jul 25, 2016

  1. CMake: link with ${HDF5_HL_LIBRARIES} 

    Fixes issue BVLC#3224.
    @intelfx
    intelfx committed Jul 25, 2016
    Configuration menu
    Copy the full SHA
    0ad1284 View commit details
    Browse the repository at this point in the history

Commits on Jul 26, 2016

  1. Fix search for Atlas on arch.

    Hans Gaiser committed Jul 26, 2016
    Configuration menu
    Copy the full SHA
    c62e06b View commit details
    Browse the repository at this point in the history

Commits on Jul 28, 2016

  1. add cudnn interfaces for n-dimensional computation

    Christian Payer committed Jul 28, 2016
    Configuration menu
    Copy the full SHA
    bc1a433 View commit details
    Browse the repository at this point in the history

  2. add support for nd convolution in cudnn

    Christian Payer committed Jul 28, 2016
    Configuration menu
    Copy the full SHA
    e5c13a5 View commit details
    Browse the repository at this point in the history

  3. change interface of pool to support n-dimensions 

    support n-dimensional pooling for cudnn
caffe cpu and gpu pooling implementations do not work in this revision!
    Christian Payer committed Jul 28, 2016
    Configuration menu
    Copy the full SHA
    cc357bd View commit details
    Browse the repository at this point in the history

  4. fix 2D pooling on CPU and GPU 

    change height and width parameters of pooling layer
to new vector interfaces
    Christian Payer committed Jul 28, 2016
    Configuration menu
    Copy the full SHA
    12cb24f View commit details
    Browse the repository at this point in the history

  5. remove some calls of Blob::LegacyShape() to support 3D

    Christian Payer committed Jul 28, 2016
    Configuration menu
    Copy the full SHA
    c1b0f38 View commit details
    Browse the repository at this point in the history

  6. fix xavier filler to use new blob shape accessors

    Christian Payer committed Jul 28, 2016
    Configuration menu
    Copy the full SHA
    721553e View commit details
    Browse the repository at this point in the history

  7. fix tests for new pooling parameter interface

    Christian Payer committed Jul 28, 2016
    Configuration menu
    Copy the full SHA
    b2f3848 View commit details
    Browse the repository at this point in the history

  8. add 3D cudnn convolution tests 

    the 2 tests copied and adapted from 3D caffe convolution tests
    Christian Payer committed Jul 28, 2016
    Configuration menu
    Copy the full SHA
    7173035 View commit details
    Browse the repository at this point in the history

  9. add 3D cudnn pooling tests 

    test the same things as the 2D pooling tests
    Christian Payer committed Jul 28, 2016
    Configuration menu
    Copy the full SHA
    c9de153 View commit details
    Browse the repository at this point in the history

  10. fix CUDNN_BAD_PARAM when using InnerProduct layer

    Christian Payer committed Jul 28, 2016
    Configuration menu
    Copy the full SHA
    eb93d32 View commit details
    Browse the repository at this point in the history

  11. change interface for cudnn v5

    Christian Payer committed Jul 28, 2016
    Configuration menu
    Copy the full SHA
    919b6d7 View commit details
    Browse the repository at this point in the history

Commits on Aug 4, 2016

  1. Merge pull request BVLC#4523 from delftrobotics/cmake-atlas 

    Fix search for Atlas on arch.
    @longjon
    longjon committed Aug 4, 2016
    Configuration menu
    Copy the full SHA
    9e9e9ba View commit details
    Browse the repository at this point in the history

  2. Merge pull request BVLC#4516 from intelfx/BVLC-work 

    CMake: link with ${HDF5_HL_LIBRARIES}
    @longjon
    longjon committed Aug 4, 2016
    Configuration menu
    Copy the full SHA
    6431477 View commit details
    Browse the repository at this point in the history

Commits on Aug 7, 2016

  1. num in blob is deprecated

    @fyu
    fyu committed Aug 7, 2016
    Configuration menu
    Copy the full SHA
    61e0165 View commit details
    Browse the repository at this point in the history

  2. Merge pull request BVLC#4559 from fyu/loss_reshape 

    Fix Label Shape Check in LossLayer::Reshape
    @jeffdonahue
    jeffdonahue committed Aug 7, 2016
    Configuration menu
    Copy the full SHA
    375003a View commit details
    Browse the repository at this point in the history

Commits on Aug 10, 2016

  1. add cudnn interfaces for n-dimensional computation

    Christian Payer committed Aug 10, 2016
    Configuration menu
    Copy the full SHA
    f86a099 View commit details
    Browse the repository at this point in the history

  2. add support for nd convolution in cudnn

    Christian Payer committed Aug 10, 2016
    Configuration menu
    Copy the full SHA
    4f63ea5 View commit details
    Browse the repository at this point in the history

  3. change interface of pool to support n-dimensions 

    support n-dimensional pooling for cudnn
caffe cpu and gpu pooling implementations do not work in this revision!
    Christian Payer committed Aug 10, 2016
    Configuration menu
    Copy the full SHA
    5e1f04e View commit details
    Browse the repository at this point in the history

  4. fix 2D pooling on CPU and GPU 

    change height and width parameters of pooling layer
to new vector interfaces
    Christian Payer committed Aug 10, 2016
    Configuration menu
    Copy the full SHA
    2346c5e View commit details
    Browse the repository at this point in the history

  5. remove some calls of Blob::LegacyShape() to support 3D

    Christian Payer committed Aug 10, 2016
    Configuration menu
    Copy the full SHA
    0dcb68a View commit details
    Browse the repository at this point in the history

  6. fix xavier filler to use new blob shape accessors

    Christian Payer committed Aug 10, 2016
    Configuration menu
    Copy the full SHA
    fb0f9f5 View commit details
    Browse the repository at this point in the history

  7. fix tests for new pooling parameter interface

    Christian Payer committed Aug 10, 2016
    Configuration menu
    Copy the full SHA
    b8ca687 View commit details
    Browse the repository at this point in the history

  8. add 3D cudnn convolution tests 

    the 2 tests copied and adapted from 3D caffe convolution tests
    Christian Payer committed Aug 10, 2016
    Configuration menu
    Copy the full SHA
    c88f8fa View commit details
    Browse the repository at this point in the history

  9. add 3D cudnn pooling tests 

    test the same things as the 2D pooling tests
    Christian Payer committed Aug 10, 2016
    Configuration menu
    Copy the full SHA
    d0efc10 View commit details
    Browse the repository at this point in the history

  10. fix CUDNN_BAD_PARAM when using InnerProduct layer

    Christian Payer committed Aug 10, 2016
    Configuration menu
    Copy the full SHA
    45562a0 View commit details
    Browse the repository at this point in the history

  11. change interface for cudnn v5

    Christian Payer committed Aug 10, 2016
    Configuration menu
    Copy the full SHA
    b506327 View commit details
    Browse the repository at this point in the history

Commits on Sep 12, 2016

  1. remove some calls of Blob::LegacyShape() to support 3D

    Christian Payer committed Sep 12, 2016
    Configuration menu
    Copy the full SHA
    fc39d7e View commit details
    Browse the repository at this point in the history

  2. fix msra filler to use new blob shape accessors

    Christian Payer committed Sep 12, 2016
    Configuration menu
    Copy the full SHA
    857f47d View commit details
    Browse the repository at this point in the history

  3. fix positive_unitball filler to use new blob shape accessors

    Christian Payer committed Sep 12, 2016
    Configuration menu
    Copy the full SHA
    334e76f View commit details
    Browse the repository at this point in the history

Commits on Nov 2, 2016

  1. Merge branch 'nd-cudnn' of github.com:christianpayer/caffe into nd-cudnn

    Christian Payer committed Nov 2, 2016
    Configuration menu
    Copy the full SHA
    efda84c View commit details
    Browse the repository at this point in the history