文章目录
CmakeLists.txt
代码语言：javascript
复制cmake_minimum_required (VERSION 3.8)
project(SOLDIER)
set(Torch_DIR "/libtorch/share/cmake/Torch")
set(PYTHON_EXECUTABLE "/usr/bin/python3")
find_package(Torch REQUIRED)
find_package(OpenCV REQUIRED)
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS}")
set(CMAKE_BUILD_TYPE Debug)
include_directories(${CMAKE_SOURCE_DIR}/include)
include_directories(${OpenCV_INCLUDE_DIRS})
add_executable(run main.cpp)
target_link_libraries(run "${OpenCV_LIBS}" "${TORCH_LIBRARIES}")
set_property(TARGET run PROPERTY CXX_STANDARD 14)
C
代码语言：javascript
复制#include <iostream>
#include<torch/torch.h>
#include<ATen/ATen.h>
#include <torch/csrc/autograd/variable.h>
#include <torch/csrc/autograd/function.h>
using namespace torch::autograd;
void basic_autograd_operations_example() {
std::cout << "====== Running: "Basic autograd operations" ======" << std::endl;
// 创建一个张量并设置为 ``torch::requires_grad()`` 用于跟踪计算过程
auto x = torch::ones({2, 2}, torch::requires_grad());
std::cout <<"x="<< x << std::endl;
// 张量的加法操作:
auto y = x 2;
std::cout << "y="<<y << std::endl;
// y是加法的记过,可以操作``grad_fn``
std::cout <<"y.grad_fn:"<< y.grad_fn()->name() << std::endl;
// 做更多的操作在y上
auto z = y * y * 3;
auto out = z.mean();
std::cout << "z="<<z << std::endl;
std::cout <<"z.grad_fn:"<< z.grad_fn()->name() << std::endl;
std::cout <<"out="<< out << std::endl;
std::cout <<"out.grad_fn:"<< out.grad_fn()->name() << std::endl;
// ``.requires_grad_( ... )`` changes an existing tensor's ``requires_grad`` flag in-place.
auto a = torch::randn({2, 2});
a = ((a * 3) / (a - 1));
std::cout << a.requires_grad() << std::endl;
a.requires_grad_(true);
std::cout << a.requires_grad() << std::endl;
auto b = (a * a).sum();
std::cout << b.grad_fn()->name() << std::endl;
// Let's backprop now. Because ``out`` contains a single scalar, ``out.backward()``
// is equivalent to ``out.backward(torch::tensor(1.))``.
out.backward();
// Print gradients d(out)/dx
std::cout << x.grad() << std::endl;
// Now let's take a look at an example of vector-Jacobian product:
x = torch::randn(3, torch::requires_grad());
y = x * 2;
while (y.norm().item<double>() < 1000) {
y = y * 2;
}
std::cout << y << std::endl;
std::cout << y.grad_fn()->name() << std::endl;
// If we want the vector-Jacobian product, pass the vector to ``backward`` as argument:
auto v = torch::tensor({0.1, 1.0, 0.0001}, torch::kFloat);
y.backward(v);
std::cout << x.grad() << std::endl;
// You can also stop autograd from tracking history on tensors that require gradients
// either by putting ``torch::NoGradGuard`` in a code block
std::cout << x.requires_grad() << std::endl;
std::cout << x.pow(2).requires_grad() << std::endl;
{
torch::NoGradGuard no_grad;
std::cout << x.pow(2).requires_grad() << std::endl;
}
// Or by using ``.detach()`` to get a new tensor with the same content but that does
// not require gradients:
std::cout << x.requires_grad() << std::endl;
y = x.detach();
std::cout << y.requires_grad() << std::endl;
std::cout << x.eq(y).all().item<bool>() << std::endl;
}
void compute_higher_order_gradients_example() {
std::cout << "====== Running "Computing higher-order gradients in C " ======" << std::endl;
// One of the applications of higher-order gradients is calculating gradient penalty.
// Let's see an example of it using ``torch::autograd::grad``:
auto model = torch::nn::Linear(4, 3);
auto input = torch::randn({3, 4}).requires_grad_(true);
auto output = model(input);
// Calculate loss
auto target = torch::randn({3, 3});
auto loss = torch::nn::MSELoss()(output, target);
// Use norm of gradients as penalty
auto grad_output = torch::ones_like(output);
auto gradient = torch::autograd::grad({output}, {input}, /*grad_outputs=*/{grad_output}, /*create_graph=*/true)[0];
auto gradient_penalty = torch::pow((gradient.norm(2, /*dim=*/1) - 1), 2).mean();
// Add gradient penalty to loss
auto combined_loss = loss gradient_penalty;
combined_loss.backward();
std::cout << input.grad() << std::endl;
}
// Inherit from Function
class LinearFunction : public Function<LinearFunction> {
public:
// Note that both forward and backward are static functions
// bias is an optional argument
static torch::Tensor forward(
AutogradContext *ctx, torch::Tensor input, torch::Tensor weight, torch::Tensor bias = torch::Tensor()) {
ctx->save_for_backward({input, weight, bias});
auto output = input.mm(weight.t());
if (bias.defined()) {
output = bias.unsqueeze(0).expand_as(output);
}
return output;
}
static tensor_list backward(AutogradContext *ctx, tensor_list grad_outputs) {
auto saved = ctx->get_saved_variables();
auto input = saved[0];
auto weight = saved[1];
auto bias = saved[2];
auto grad_output = grad_outputs[0];
auto grad_input = grad_output.mm(weight);
auto grad_weight = grad_output.t().mm(input);
auto grad_bias = torch::Tensor();
if (bias.defined()) {
grad_bias = grad_output.sum(0);
}
return {grad_input, grad_weight, grad_bias};
}
};
class MulConstant : public Function<MulConstant> {
public:
static torch::Tensor forward(AutogradContext *ctx, torch::Tensor tensor, double constant) {
// ctx is a context object that can be used to stash information
// for backward computation
ctx->saved_data["constant"] = constant;
return tensor * constant;
}
static tensor_list backward(AutogradContext *ctx, tensor_list grad_outputs) {
// We return as many input gradients as there were arguments.
// Gradients of non-tensor arguments to forward must be `torch::Tensor()`.
return {grad_outputs[0] * ctx->saved_data["constant"].toDouble(), torch::Tensor()};
}
};
void custom_autograd_function_example() {
std::cout << "====== Running "Using custom autograd function in C " ======" << std::endl;
{
auto x = torch::randn({2, 3}).requires_grad_();
auto weight = torch::randn({4, 3}).requires_grad_();
auto y = LinearFunction::apply(x, weight);
y.sum().backward();
std::cout << x.grad() << std::endl;
std::cout << weight.grad() << std::endl;
}
{
auto x = torch::randn({2}).requires_grad_();
auto y = MulConstant::apply(x, 5.5);
y.sum().backward();
std::cout << x.grad() << std::endl;
}
}
int autograde_example(){
std::cout << "Autograd: Automatic Differentiationnn";
// Create a tensor and set requires_grad=True to track computation with it:
auto x = torch::ones({2, 2}, torch::TensorOptions().requires_grad(true));
std::cout << "x:n" << x << 'n';
// Do a tensor operation:
auto y = x 2;
std::cout << "y:n" << y << 'n';
// y was created as a result of an operation, so it has a grad_fn:
std::cout << "y.grad_fn:n" << y.grad_fn() << 'n';
// Do more operations on y:
auto z = y * y * 3;
auto out = z.mean();
std::cout << "z:n" << z << "out:n" << out << 'n';
// .requires_grad_(...) changes an existing Tensor’s requires_grad flag in-place:
auto a = torch::randn({2, 2});
a = ((a * 3) / (a - 1));
std::cout << a.requires_grad() << 'n';
a.requires_grad_(true);
std::cout << a.requires_grad() << 'n';
auto b = (a * a).sum();
std::cout << b.grad_fn() << 'n';
std::cout << "Gradientsnn";
// Let’s backprop now:
out.backward();
// Print gradients d(out)/dx:
std::cout << "x.grad:n" << x.grad() << 'n';
// Example of vector-Jacobian product:
x = torch::randn(3, torch::TensorOptions().requires_grad(true));
y = x * 2;
while (y.data().norm().item<int>() < 1000) {
y = y * 2;
}
std::cout << "y:n" << y << 'n';
// Simply pass the vector to backward as argument:
auto v = torch::tensor({0.1, 1.0, 0.0001}, torch::TensorOptions(torch::kFloat));
y.backward(v);
std::cout << "x.grad:n" << x.grad() << 'n';
// Stop autograd from tracking history on Tensors with .requires_grad=True:
std::cout << "x.requires_gradn" << x.requires_grad() << 'n';
std::cout << "(x ** 2).requires_gradn" << (x * x).requires_grad() << 'n';
torch::NoGradGuard no_grad;
std::cout << "(x ** 2).requires_gradn" << (x * x).requires_grad() << 'n';
// Or by using .detach() to get a new Tensor with the same content but that does not require gradients:
std::cout << "x.requires_grad:n" << x.requires_grad() << 'n';
y = x.detach();
std::cout << "y.requires_grad:n" << y.requires_grad() << 'n';
std::cout << "x.eq(y).all():n" << x.eq(y).all() << 'n';
return 0;
}
int main()
{
std::cout << "Deep Learning with PyTorch" << std::endl;
basic_autograd_operations_example();
std::cout << "n";
compute_higher_order_gradients_example();
std::cout << "n";
custom_autograd_function_example();
std::cout<< "n";
autograde_example();
return 0;
}
