名称 最后提交 最后更新
..
CMakeLists.txt Add AXPY oneDNN handler (#33632)
activation_mkldnn_op.cc [oneDNN] Fix to 34554 (same as previous PR but should build with GPU) (#34859)
axpy_handler.cc [OneDNN] Conv op refactor. (#36252)
axpy_handler.h Reuse OneDNN handler for SGD and SUM for SelectedRows input tensors. (#35510)
batch_norm_mkldnn_op.cc fix bn/in/squeeze/syncbn extra (#35502)
caching_tests.cmake [oneDNN] Fix to 34554 (same as previous PR but should build with GPU) (#34859)
cast_mkldnn_op.cc [oneDNN] Disable caching of Reorder operation (#35664)
clip_mkldnn_op.cc Added clip BF16/FP32 FWD/BWD kernels (#35601)
concat_mkldnn_op.cc Added concat BF16/FP32 BWD OneDNN kernel (#35889)
conv_mkldnn_op.cc [OneDNN] Conv op refactor. (#36252)
conv_transpose_mkldnn_op.cc Add Conv Transpose BF16 (#30877)
dequantize_mkldnn_op.cc [oneDNN] Cache oneDNN stream not to recreate in each oneDNN op (#30358)
expand_v2_mkldnn_op.cc [oneDNN] Disable caching of Reorder operation (#35664)
fc_mkldnn_op.cc Copy boost optional to Paddle (#34780)
gaussian_random_mkldnn_op.cc add empty op (c++, python, unit test) (#26659)
inplace_op_tests.cmake [DNNL] activations Inplace support (#24123)
interpolate_mkldnn_op.cc [oneDNN] disable caching for interpolate and batch Norm (#35030)
layer_norm_mkldnn_op.cc [oneDNN ] disabling more ops caching (#34830)
lrn_mkldnn_op.cc [oneDNN ] disabling more ops caching (#34830)
matmul_mkldnn_op.cc [oneDNN] Disable caching of Reorder operation (#35664)
matmul_mkldnn_op.h [oneDNN] disable caching oneDNN primitives in matmul v2, Reduce grad and elementwise_add grad, expand_v2 (#35132)
matmul_v2_mkldnn_op.cc [oneDNN] disable caching oneDNN primitives in matmul v2, Reduce grad and elementwise_add grad, expand_v2 (#35132)
mkldnn_activation_op.h Update PADDLE_ENFORCE in DNNL related ops (#24333)
mul_mkldnn_op.cc Copy boost optional to Paddle (#34780)
nhwc_op_tests.cmake Fix to issue #25537 (#27546)
pool_mkldnn_op.cc
prelu_mkldnn_op.cc
quantize_mkldnn_op.cc
requantize_mkldnn_op.cc
reshape_mkldnn_op.cc
scale_mkldnn_op.cc
slice_mkldnn_op.cc
softmax_mkldnn_op.cc
split_mkldnn_op.cc
sum_mkldnn_op.cc
test_mkldnn_caching.cc
test_mkldnn_op_inplace.cc
test_mkldnn_op_nhwc.cc
transpose_mkldnn_op.cc

项目简介

PArallel Distributed Deep LEarning: Machine Learning Framework from Industrial Practice (『飞桨』核心框架,深度学习&机器学习高性能单机、分布式训练和跨平台部署)

:rocket: Github 镜像仓库 :rocket:

源项目地址 :arrow_down: :arrow_down: :arrow_down:

https://github.com/paddlepaddle/paddle

发行版本

当前项目没有发行版本

贡献者 228

全部贡献者

开发语言

  • C++ 45.8 %
  • Python 45.5 %
  • Cuda 6.4 %
  • CMake 1.1 %
  • Shell 0.7 %
反馈
建议
客服 返回
顶部