Commit Graph

255 Commits (b007055e9d72fc8cb00177aa89cc4fbb245ef8b2)

Author SHA1 Message Date
Yi Wang 4558807c48 Use PADDLE_WITH_CUDA instead of PADDLE_WITH_GPU
7 years ago
Yu Yang 84500f9487 Change `PADDLE_ONLY_CPU` to `PADDLE_WITH_GPU`
7 years ago
Liu Yiqun 3fefee8a06 Use scalar implementation instead of neon implementation to avoid out of range memory access in the tail conv3x3.
7 years ago
Liu Yiqun 9e524fce1b Remove the generation of executable, such as paddle_trainer, when setting WITH_C_API.
8 years ago
hedaoyuan 0146e8bfc5 Merge pull request #3843 from Xreki/build_android_clang
8 years ago
hedaoyuan a8efed0910 Merge pull request #3792 from hedaoyuan/convolution
8 years ago
Liu Yiqun 3360e9cdb8 Change the definition of vmlaq_laneq_f32 from template function to macro.
8 years ago
wanghaoshuang e033569dd6 Fix format
8 years ago
wanghaoshuang eb3c774b83 Fix format error
8 years ago
whs a162ebce8b Merge branch 'develop' into pixel_softmax_layer
8 years ago
Liu Yiqun 387b4e4eac Change the declaration of EigenGemm from `class` to `struct`.
8 years ago
Liu Yiqun 8b15ac82fa Move the definition of hl_cpu_gru_forward and hl_cpu_gru_backward to function/GruFunctor.h.
8 years ago
Liu Yiqun c54c7d91a0 Use template to deliver const argument instead, to remove the compiling error "argument to __builtin_neon_vgetq_lane_f32 must be a constant integer".
8 years ago
Liu Yiqun 50c63dc9e7 Merge branch 'develop' into build_android_clang
8 years ago
hedaoyuan 90bf4f60ae Add stride support 2 for NeonDepthwiseConvTranspose.
8 years ago
hedaoyuan 840104c99a Add NeonDepthwiseConvTransposeFunction.
8 years ago
hedaoyuan 40d47fae95 [Refine code]Move class Padding into the NeonDepthwiseConv.h.
8 years ago
hedaoyuan 4b6b7251c1 Refine NeonDepthwiseConv.
8 years ago
hedaoyuan f7e75a03cf Refine the neon depthwise convolution code(separate the Function and kernel).
8 years ago
hedaoyuan 5df384d67f Remove NeonDepthwiseConv.h
8 years ago
hedaoyuan 3a75b4b70c Fix CMakeLists.text
8 years ago
hedaoyuan 227fdfb65d Refine NeonDepthwiseConvFunction.
8 years ago
Liu Yiqun f241773c4f Support to use clang for Android cross-compiling.
8 years ago
hedaoyuan f00c4112d2 Neon depthwise conv with filterSize = 4 and stride = 2.
8 years ago
hedaoyuan 6dcff9a444 Neon depthwise conv with filterSize = 3 and stride = 2.
8 years ago
hedaoyuan b7885b087b Add DepthwiseConvKernel for filter size is 4.
8 years ago
hedaoyuan 0dffe68ca9 Add NeonDepthwiseConvFunction.
8 years ago
guosheng 117ce4cbc1 Change class to struct in GemmFunctor to avoid errors on special compilers
8 years ago
hedaoyuan 430e0e418b Follow comments.
8 years ago
hedaoyuan 6ba04dcd11 Remove the header files that do not need to be included.
8 years ago
hedaoyuan adcca2cc06 Add PADDLE_USE_EIGEN_FOR_BLAS macro.
8 years ago
hedaoyuan ec2ba24206 Fix GemmConvFunction.
8 years ago
hedaoyuan 53b0e42709 Add EigenGemm.
8 years ago
hedaoyuan da616a6f2f Fix some bugs.
8 years ago
hedaoyuan 370dcf7654 Add a todo in the nnpack function, when groups > 1 and batchSize > 1.
8 years ago
hedaoyuan f4bb60ae37 Refine NNPACKConvOpTest.
8 years ago
hedaoyuan 1d74d16cca Remove the useless code.
8 years ago
hedaoyuan 03799bdbfe Refine the unit test of convolution function.
8 years ago
Yi Wang 6c66044ce0 Fix more warnings about comparison between signed and unsigned values
8 years ago
Yi Wang 4480115681 Fix some build warnings
8 years ago
qingqing01 455ad5b5b2 Merge pull request #3141 from hedaoyuan/nnpack
8 years ago
Yi Wang fe04175ed3 Fix comparison between signed and unsigned values
8 years ago
gangliao 75185d821a Merge pull request #3228 from gangliao/clang-format
8 years ago
hedaoyuan d18a9f1f37 Simplify BlockExpandOpTest.
8 years ago
hedaoyuan 7b55a0cf14 Simplify CrossMapNormalOpTest.
8 years ago
liaogang 1d4fa2433c ClangFormat for proto and cuda
8 years ago
xzl fa10677aed modify skipIm2col to need2col, delete useless variable colBuffer
8 years ago
xzl 5229df52a5 ignore im2col if not necessary in conv 1 * 1
8 years ago
hedaoyuan 953f8ddf05 Support groups in NNPACKFunction.
8 years ago
hedaoyuan ff8262e283 Merge branch 'develop' into ImageExpandFunction
8 years ago