Struct emulator::inference::LibTorchBackend::Impl
ClassList > emulator > inference > LibTorchBackend > Impl
Private implementation details for LibTorchBackend .More...
Public Attributes
| Type | Name |
|---|---|
| torch::Device | device = torch::kCPUExecution device (CPU or CUDA) |
| torch::ScalarType | dtype = torch::kFloat32Model precision. |
| torch::jit::script::Module | model Loaded TorchScript model. |
| bool | model_loaded = falseWhether model was successfully loaded. |
Detailed Description
Uses PIMPL idiom to hide LibTorch types from the header, avoiding the need to include torch headers in dependent code.
Public Attributes Documentation
variable device
Execution device (CPU or CUDA)
torch::Device emulator::inference::LibTorchBackend::Impl::device;
variable dtype
Model precision.
torch::ScalarType emulator::inference::LibTorchBackend::Impl::dtype;
variable model
Loaded TorchScript model.
torch::jit::script::Module emulator::inference::LibTorchBackend::Impl::model;
variable model_loaded
Whether model was successfully loaded.
bool emulator::inference::LibTorchBackend::Impl::model_loaded;
The documentation for this class was generated from the following file components/emulator_comps/common/src/inference/libtorch_backend.cpp