chainer.backend.ChainerxDevice¶
- class chainer.backend.ChainerxDevice(device: chainerx.Device)[source]¶
Device for ChainerX backend
Methods
- __enter__()[source]¶
A dummy definition that simply raises RuntimeError.
chainer.using_device()
should be used instead.
- create_context()[source]¶
Returns a context manager in which the device is made current.
See also
chainer.using_device()
calls this method internally.
- static from_fallback_device(device)[source]¶
Returns a
ChainerxDevice
corresponding to the fallback device.See also
- is_array_supported(array)[source]¶
Returns if the specified array is compatible with the device. :param array: An array to be checked :type array: N-dimensional array
- Returns
True
if the array is compatible with the device. OtherwiseFalse
is returned.
- send(arrays)[source]¶
Transfers given arrays to the device.
- Parameters
arrays – Array or arrays of NumPy, CuPy, or ChainerX.
- Returns
Transferred arrays.
- __lt__(value, /)¶
Return self<value.
- __le__(value, /)¶
Return self<=value.
- __gt__(value, /)¶
Return self>value.
- __ge__(value, /)¶
Return self>=value.
Attributes
- fallback_device¶
Fallback device.
A fallback device is either a
CpuDevice
or aGpuDevice
which shares the same physical device with the original ChainerX device.For example, the fallback device of
native:0
ChainerX device isCpuDevice
. The fallback device ofcuda:1
ChainerX device isGpuDevice
with device ID 1.
- name¶
- supported_array_types = (<class 'chainerx.ndarray'>,)¶