neon.layers.layer.Conv

class neon.layers.layer.Conv(fshape, init, strides={}, padding={}, dilation={}, bias=None, batch_norm=False, activation=None, name=None)[source]

Bases: neon.layers.layer.CompoundLayer

A convolutional layer with a learned bias and activation, implemented as a list composing separate Convolution, Bias and Activation layers.

Parameters:
  • fshape (tuple(int)) – three dimensional shape of convolution window
  • init (Initializer, optional) – Initializer object to use for initializing layer weights
  • strides (int, dict, optional) – strides to apply convolution window over. An int applies to all dimensions, or a dict with str_h and str_w applies to h and w dimensions distinctly. Defaults to str_w = str_h = 1
  • padding (int, dict, optional) – padding to apply to edges of input. An int applies to all dimensions, or a dict with pad_h and pad_w applies to h and w dimensions distinctly. Defaults to pad_w = pad_h = 0
  • dilation (int, dict, optional) – dilation to apply to dimensions of the filter. An int applies to all dimensions, or a dict with dil_h and dil_w applies to h and w dimensions distinctly. Defaults to dil_w = dil_h = 1
  • bias (Initializer) – an initializer to use for bias parameters
  • activation (Transform) – a transform object with fprop and bprop functions to apply
  • name (str) – the root name for the layer, suffixes are automatically generated for the component layers
__init__(fshape, init, strides={}, padding={}, dilation={}, bias=None, batch_norm=False, activation=None, name=None)[source]

Methods

__init__(fshape, init[, strides, padding, …])
add_postfilter_layers()
append((object) -> None – append object to end)
clear(() -> None – remove all items from L)
copy(() -> list – a shallow copy of L)
count(…)
extend(…)
index((value, [start, …) Raises ValueError if the value is not present.
init_base_name()
insert L.insert(index, object) – insert object before index
pop(…) Raises IndexError if list is empty or index is out of range.
remove(…) Raises ValueError if the value is not present.
reverse L.reverse() – reverse IN PLACE
sort((key=None[, reverse])
add_postfilter_layers()[source]
append(object) → None -- append object to end
clear() → None -- remove all items from L
copy() → list -- a shallow copy of L
count(value) → integer -- return number of occurrences of value
extend(iterable) → None -- extend list by appending elements from the iterable
index(value[, start[, stop]]) → integer -- return first index of value.

Raises ValueError if the value is not present.

init_base_name()
insert()

L.insert(index, object) – insert object before index

pop([index]) → item -- remove and return item at index (default last).

Raises IndexError if list is empty or index is out of range.

remove(value) → None -- remove first occurrence of value.

Raises ValueError if the value is not present.

reverse()

L.reverse() – reverse IN PLACE

sort(key=None, reverse=False) → None -- stable sort *IN PLACE*