Commits

Breadcrumbs

History for

ondevice-abstraction

User selector

All users

All time

Commit History

Commits on Aug 12, 2014

• Fix post-rebase errors, including:

  Show description for d52d669
  jeffdonahue

  committed
  d52d669

  Copy full SHA for d52d669

• Fix post-rebase errors.

  jeffdonahue

  committed
  e30ba4d

  Copy full SHA for e30ba4d

• Add "#ifndef CPU_ONLY" guards around GPU device code so that CPU_ONLY

  Show description for afd7bf5
  jeffdonahue

  committed
  afd7bf5

  Copy full SHA for afd7bf5

• include math.h in filler and solver -- seems to be needed for pow and

  Show description for 95bf45f
  jeffdonahue

  committed
  95bf45f

  Copy full SHA for 95bf45f

• Fix all rebase errors. All tests are passing.

  

  Rob Hess

  authored and

  jeffdonahue

  committed
  9a5b6d2

  Copy full SHA for 9a5b6d2

• Use UVA for GPU copy. Get rid of copy_from_cpu().

  

  Rob Hess

  authored and

  jeffdonahue

  committed
  e68a140

  Copy full SHA for e68a140

• Separate device.hpp into device-specific headers.

  

  Rob Hess

  authored and

  jeffdonahue

  committed
  39c8658

  Copy full SHA for 39c8658

• Implement all kernel device functions under GPUDevice. Get rid of unused files.

  

  Rob Hess

  authored and

  jeffdonahue

  committed
  c125a97

  Copy full SHA for c125a97

• Implement all non-kernel gpu functions in GPUDevice and replace corresponding calls.

  Show description for e1c66dc
  

  Rob Hess

  authored and

  jeffdonahue

  committed
  e1c66dc

  Copy full SHA for e1c66dc

• Move caffe_rng_rand() into util/rng.hpp.

  

  Rob Hess

  authored and

  jeffdonahue

  committed
  af5e292

  Copy full SHA for af5e292

• Get rid of all `caffe_*` cpu functions.

  

  Rob Hess

  authored and

  jeffdonahue

  committed
  8fb944c

  Copy full SHA for 8fb944c

• Replace all `caffe_*()` cpu functions with `CPUDevice` functions.

  

  Rob Hess

  authored and

  jeffdonahue

  committed
  f721ac6

  Copy full SHA for f721ac6

• Try to obey google style guide.

  

  Rob Hess

  authored and

  jeffdonahue

  committed
  bddcae8

  Copy full SHA for bddcae8

• Fix formatting.

  

  Rob Hess

  authored and

  jeffdonahue

  committed
  9420cea

  Copy full SHA for 9420cea

• Fix lint errors.

  

  Rob Hess

  authored and

  jeffdonahue

  committed
  72b0af1

  Copy full SHA for 72b0af1

• Fix compilation errors.

  

  Rob Hess

  authored and

  jeffdonahue

  committed
  1393aa4

  Copy full SHA for 1393aa4

• Incorporate all cpu math functions directly into CPUDevice.

  

  Rob Hess

  authored and

  jeffdonahue

  committed
  6db9c8c

  Copy full SHA for 6db9c8c

• Make device factory usage more inline with layer factory.

  

  Rob Hess

  authored and

  jeffdonahue

  committed
  96ed3c2

  Copy full SHA for 96ed3c2

• Move device.hpp up out of util/.

  

  Rob Hess

  authored and

  jeffdonahue

  committed
  ff9bff0

  Copy full SHA for ff9bff0

• Incorporate additional abstractions from 1605431.

  

  Rob Hess

  authored and

  jeffdonahue

  committed
  d45100f

  Copy full SHA for d45100f

• Fix rebase errors.

  

  Rob Hess

  authored and

  jeffdonahue

  committed
  99380c9

  Copy full SHA for 99380c9

• Add current-device data accessors to SyncedMem. Use to fix ConvolutionLayer.

  Show description for 6618265
  

  Rob Hess

  authored and

  jeffdonahue

  committed
  6618265

  Copy full SHA for 6618265

• Device wrapper methods no longer pure virtual, default not implemented

  Show description for 56e23fa
  

  kloudkl

  authored and

  jeffdonahue

  committed
  56e23fa

  Copy full SHA for 56e23fa

• Update the include guard of the util/device.hpp

  

  kloudkl

  authored and

  jeffdonahue

  committed
  31fb4b9

  Copy full SHA for 31fb4b9

• Move im2col and col2im into the device wrapper classes

  

  kloudkl

  authored and

  jeffdonahue

  committed
  4890b55

  Copy full SHA for 4890b55

• Unify the CPU/GPU versions of Forward/Backward of the PowerLayer

  

  kloudkl

  authored and

  jeffdonahue

  committed
  ae17c17

  Copy full SHA for ae17c17

• Unify the CPU/GPU versions of Forward/Backward of the Im2colLayer

  

  kloudkl

  authored and

  jeffdonahue

  committed
  f70db65

  Copy full SHA for f70db65

• Unify the CPU/GPU versions of Forward/Backward of the ConvolutionLayer

  

  kloudkl

  authored and

  jeffdonahue

  committed
  0275bdf

  Copy full SHA for 0275bdf

• Add im2col and col2im to wrap im2col_{cpu, gpu} and col2im_{cpu, gpu}

  

  kloudkl

  authored and

  jeffdonahue

  committed
  740820b

  Copy full SHA for 740820b

• Unify the CPU/GPU Forward/Backward of the EltwiseLayer

  

  kloudkl

  authored and

  jeffdonahue

  committed
  faef386

  Copy full SHA for faef386

• Unify the CPU/GPU Forward/Backward of the InnerProductLayer

  

  kloudkl

  authored and

  jeffdonahue

  committed
  716c29d

  Copy full SHA for 716c29d

• Consolidate the CPU and GPU Forward of the WindowDataLayer

  

  kloudkl

  authored and

  jeffdonahue

  committed
  9b01ff1

  Copy full SHA for 9b01ff1

• Merge the CPU and the GPU Backward of the data layers

  

  kloudkl

  authored and

  jeffdonahue

  committed
  ed7582b

  Copy full SHA for ed7582b

• Unify the CPU and the GPU Forward & Backward of the HDF5OutputDataLayer

  

  kloudkl

  authored and

  jeffdonahue

  committed
  0703d7b

  Copy full SHA for 0703d7b

• Unify the CPU and the GPU Forward of the ImageDataLayer

  

  kloudkl

  authored and

  jeffdonahue

  committed
  e57a703

  Copy full SHA for e57a703