pytorch  1.8.2
About: PyTorch provides Tensor computation (like NumPy) with strong GPU acceleration and Deep Neural Networks (in Python) built on a tape-based autograd system. LTS (Long Term Support) release.
  Fossies Dox: pytorch-1.8.2.tar.gz  ("unofficial" and yet experimental doxygen-generated source code documentation)  

bench_approx.cpp File Reference
Include dependency graph for bench_approx.cpp:

Go to the source code of this file.


void optimizeSleef (tensorexpr::LoopNest *ln, tensorexpr::Tensor *target)
void optimizePointwise (tensorexpr::LoopNest *ln, tensorexpr::Tensor *target)
static void log_nnc_sleef (benchmark::State &state)
static void log_nnc_fast (benchmark::State &state)
static void log_aten (benchmark::State &state)
static void logit_nnc_sleef (benchmark::State &state)
static void logit_nnc_fast (benchmark::State &state)
static void logit_aten (benchmark::State &state)
template<typename T >
void logit_caffe2_impl (int size, const T *X, T *Y, float eps_=1e-6f)
static void logit_caffe2 (benchmark::State &state)
static void tanh_nnc_fast (benchmark::State &state)
static void tanh_aten (benchmark::State &state)
static void tanh_caffe2 (benchmark::State &state)
 Args ({2<< 5}) -> Args({2<< 8}) ->Args({2<< 12}) ->Args({2<< 14})

Function Documentation

◆ Args()

◆ log_aten()

static void log_aten ( benchmark::State &  state)

◆ log_nnc_fast()

◆ log_nnc_sleef()

◆ logit_aten()

static void logit_aten ( benchmark::State &  state)

◆ logit_caffe2()

◆ logit_caffe2_impl()

template<typename T >
void logit_caffe2_impl ( int  size,
const T *  X,
T *  Y,
float  eps_ = 1e-6f 

Definition at line 186 of file bench_approx.cpp.

References size, X, and at::native::metal::mpscnn::Y.

Referenced by logit_caffe2().

◆ logit_nnc_fast()

◆ logit_nnc_sleef()

◆ optimizePointwise()

◆ optimizeSleef()

◆ tanh_aten()

static void tanh_aten ( benchmark::State &  state)

◆ tanh_caffe2()

◆ tanh_nnc_fast()