Package | Description |
---|---|
org.bytedeco.pytorch.cuda | |
org.bytedeco.pytorch.global |
Modifier and Type | Method and Description |
---|---|
CUDAStream |
CUDAStreamArrayRef.at(long Index)
Vector compatibility
|
CUDAStream |
CUDAStreamArrayRef.back()
back - Get the last element.
|
CUDAStream |
CUDAStreamArrayRef.begin()
\}
\name Simple Operations
\{
|
CUDAStream |
CUDAStreamArrayRef.cbegin() |
CUDAStream |
CUDAStreamArrayRef.cend() |
CUDAStream |
CUDAStreamGuard.current_stream()
Returns the most recent CUDA stream that was set using this device guard,
either from construction, or via set_stream.
|
CUDAStream |
CUDAStreamArrayRef.data() |
CUDAStream |
CUDAStreamArrayRef.end() |
CUDAStream |
CUDAStreamArrayRef.front()
front - Get the first element.
|
CUDAStream |
CUDAStreamArrayRef.get(long Index)
\}
\name Operator Overloads
\{
|
CUDAStream |
CUDAStreamGuard.original_stream()
Returns the CUDA stream that was set at the time the guard was
constructed.
|
static CUDAStream |
CUDAStream.unpack3(long stream_id,
byte device_index,
torch.DeviceType device_type) |
CUDAStream |
CUDAStreamArrayRef.vec()
\}
\name Expensive Operations
\{
|
Modifier and Type | Method and Description |
---|---|
boolean |
CUDAStream.equals(CUDAStream other) |
boolean |
CUDAStream.notEquals(CUDAStream other) |
void |
CUDAAllocator.recordStream(DataPtr arg0,
CUDAStream stream) |
Constructor and Description |
---|
CUDAStreamArrayRef(CUDAStream begin,
CUDAStream end)
Construct an ArrayRef from a range.
|
CUDAStreamArrayRef(CUDAStream data,
long length)
Construct an ArrayRef from a pointer and length.
|
Modifier and Type | Method and Description |
---|---|
static CUDAStream |
torch_cuda.getCurrentCUDAStream() |
static CUDAStream |
torch_cuda.getCurrentCUDAStream(byte device_index)
Get the current CUDA stream, for the passed CUDA device, or for the
current device if no device index is passed.
|
static CUDAStream |
torch_cuda.getDefaultCUDAStream() |
static CUDAStream |
torch_cuda.getDefaultCUDAStream(byte device_index)
Get the default CUDA stream, for the passed CUDA device, or for the
current device if no device index is passed.
|
static CUDAStream |
torch_cuda.getStreamFromExternal(Pointer ext_stream,
byte device_index)
Get a CUDAStream from a externally allocated one.
|
static CUDAStream |
torch_cuda.getStreamFromPool() |
static CUDAStream |
torch_cuda.getStreamFromPool(boolean isHighPriority,
byte device)
Get a new stream from the CUDA stream pool.
|
static CUDAStream |
torch_cuda.getStreamFromPool(int priority) |
static CUDAStream |
torch_cuda.getStreamFromPool(int priority,
byte device) |
Modifier and Type | Method and Description |
---|---|
static void |
torch_cuda.setCurrentCUDAStream(CUDAStream stream)
Set the current stream on the device of the passed in stream to be
the passed in stream.
|
static Pointer |
torch_cuda.shiftLeft(Pointer stream,
CUDAStream s) |
Copyright © 2024. All rights reserved.