Pinned Loading

1. AITemplateAITemplatePublic

   Forked from facebookincubator/AITemplate

   AITemplate is a Python framework which renders neural network into high performance CUDA/HIP C++ code. Specialized for FP16 TensorCore (NVIDIA GPU) and MatrixCore (AMD GPU) inference.

   Python

2. c++ multiprocessing Barrier/CondVar/...c++ multiprocessing Barrier/CondVar/SharedMemory using SysV IPC

   1

   #include"ipc.hpp"

   2

   #include<cassert>

   3

   #include<cstring>

   4

   #include<sys/ipc.h>

   5

   #include<sys/sem.h>

3. flat_array_t and flat_tuple_t are au...flat_array_t and flat_tuple_t are automatically hashable type wrapper for std::array<T, N> and std::tuple<Ts...> that can be used as keys in map/unordered_map

   1

   #include<array>

   2

   #include<functional>

   3

   #include<iostream>

   4

   #include<tuple>

   5

   #include<unordered_map>

4. host_vector (a std::vector with a cu...host_vector (a std::vector with a custom allocator based on cudaMallocHost)

   1

   #include"host_vector.hpp"

   2

   #defineDEVICE_CHECK(call) \

   3

   if ((call) != cudaSuccess) { \

   4

   throwstd::runtime_error(#call " API call failed: " + GetLastErrorString() + " at " + \

   5

    __FILE__ + ", line" + std::to_string(__LINE__)); \

5. pdf_croppdf_cropPublic

   Python

6. dlp2019dlp2019Public archive

   Deep Learning and Practice 2019

   Python