Modules
Dataset iterators.
Indexing operations
JIT interface to run model trained/saved using PyTorch Python API.
The different kind of elements supported in Torch.
A small neural-network library based on Torch.
The
vision module groups functions and models related to
computer vision.Structs
A jit PyTorch module.
A RAII guard that prevents gradient tracking until deallocated.
A single scalar value.
A tensor object.
The trainable version of a jit PyTorch module.
Enums
Cuda related helper functions.
A torch device.
Argument and output values for JIT models.
The different kind of elements that a Tensor can hold.
Quantization engines
Main library error type.
Traits
Functions
Runs a closure in mixed precision.
Get the number of threads used by torch for inter-op parallelism.
Get the number of threads used by torch in parallel regions.
Sets the random seed used by torch.
Runs a closure without keeping track of gradients.
Disables gradient tracking, this will be enabled back when the
returned value gets deallocated.
Note that it is important to bind this to a name like “guard”
and not to “” as these two would have different semantics.
See https://internals.rust-lang.org/t/pre-rfc-must-bind/12658/46
for more details.
Set the number of threads used by torch for inter-op parallelism.
Set the number of threads used by torch in parallel regions.
Runs a closure explicitly keeping track of gradients, this could be
run within a no_grad closure for example.