Package | Description |
---|---|
org.bytedeco.pytorch.cuda | |
org.bytedeco.pytorch.global |
Class and Description |
---|
ActivationDescriptor |
AllocatorState |
BlockInfo |
CheckpointDelta |
Constant |
ConvolutionDescriptor |
CTCLossDescriptor |
CUDAKernelLaunchInfo
Use to hold info about kernel launches so that we can run kernels
asynchronously and still associate launches with device-side
assertion failures
|
CUDAKernelLaunchInfoVector |
CUDAKernelLaunchInfoVector.Iterator |
CUDAKernelLaunchRegistry
Circular buffer used to hold information about kernel launches
this is later used to reconstruct how a device-side kernel assertion failure
occurred CUDAKernelLaunchRegistry is used as a singleton
|
CUDAStream |
CUDAStream.Unchecked |
CUDAStreamArrayRef
ArrayRef - Represent a constant reference to an array (0 or more elements
consecutively in memory), i.e.
|
DeviceAssertionData
Holds information about any device-side assertions that fail.
|
DeviceAssertionsData
Used to hold assertions generated by the device
Held in managed memory and access by both the CPU and the GPU.
|
DeviceAssertionsDataVector |
DeviceAssertionsDataVector.Iterator |
DeviceAssertionsDataVectorCUDAKernelLaunchInfoVectorPair |
DeviceStats |
DropoutDescriptor |
FilterDescriptor |
PointerSet |
PointerSet.Iterator |
RNNDescriptor |
SegmentInfo |
SnapshotInfo |
SpatialTransformerDescriptor |
Stat |
TensorDescriptor |
TraceEntry |
TraceEntry.Action |
TraceEntryVector |
TraceEntryVector.Iterator |
Class and Description |
---|
CUDAAllocator |
CUDAStream |
Copyright © 2024. All rights reserved.