本文整理汇总了Python中quagga.connector.Connector.assign_sequential_mean_pooling方法的典型用法代码示例。如果您正苦于以下问题:Python Connector.assign_sequential_mean_pooling方法的具体用法?Python Connector.assign_sequential_mean_pooling怎么用?Python Connector.assign_sequential_mean_pooling使用的例子?那么, 这里精选的方法代码示例或许可以为您提供帮助。您也可以进一步了解该方法所在类quagga.connector.Connector
的用法示例。
在下文中一共展示了Connector.assign_sequential_mean_pooling方法的1个代码示例,这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞,您的评价将有助于系统推荐出更棒的Python代码示例。
示例1: SequentialMeanPoolingBlock
# 需要导入模块: from quagga.connector import Connector [as 别名]
# 或者: from quagga.connector.Connector import assign_sequential_mean_pooling [as 别名]
class SequentialMeanPoolingBlock(object):
# TODO(sergii): change sequentially_tile to add_sequentially_tile, because can erase gradients
def __init__(self, matrices, device_id=None):
self.context = Context(device_id)
device_id = self.context.device_id
self.output = Matrix.empty_like(matrices[0], device_id)
learning = matrices[0].bpropagable
self.output = Connector(self.output, device_id if learning else None)
if learning:
self.matrices, self.dL_dmatrices = izip(*matrices.register_usage(device_id, device_id))
else:
self.matrices = matrices.register_usage(device_id)
self.length = matrices.length
def fprop(self):
self.output.assign_sequential_mean_pooling(self.context, self.matrices[:self.length])
self.output.fprop()
def bprop(self):
dL_doutput = self.output.backward_matrix
dL_doutput.scale(self.context, ct.c_float(1.0 / self.length))
Matrix.sequentially_tile(self.context, dL_doutput, self.dL_dmatrices[:self.length])