ZigZag - Deep Learning Hardware Design Space Exploration
This repository presents the novel version of our tried-and-tested hardware Architecture-Mapping Design Space Exploration (DSE) Framework for Deep Learning (DL) accelerators. ZigZag bridges the gap between algorithmic DL decisions and their acceleration cost on specialized accelerators through a fast and accurate hardware cost estimation.
|
Represents a user-defined temporal ordering. More...
Public Member Functions | |
def | __init__ (self, list[list[str|UnrollFactorInt]] data) |
data will look like: [['K', 12], ['C', 3]] More... | |
def | is_empty (self) |
def | is_complete (self, dict[LayerDim, UnrollFactor] temporal_loop_sizes) |
Return wether this temporal ordering matches the given, mandatory loop sizes. More... | |
def | remove_invalid_layer_dims (self, LayerDimSizes layer_dim_sizes, str layer_name="") |
def | to_legacy_format (self) |
list[PermutationConstraint] | get_constraints (self) |
def | __hash__ (self) |
![]() | |
def | __init__ (self, Any data) |
int | __len__ (self) |
Iterator[Any] | __iter__ (self) |
def | __getitem__ (self, Any key) |
bool | __contains__ (self, Any key) |
def | __str__ (self) |
def | __repr__ (self) |
Any | __jsonrepr__ (self) |
def | __eq__ (self, object other) |
Static Public Member Functions | |
def | empty () |
Public Attributes | |
data | |
![]() | |
data | |
Represents a user-defined temporal ordering.
def __init__ | ( | self, | |
list[list[str | UnrollFactorInt]] | data | ||
) |
data will look like: [['K', 12], ['C', 3]]
def __hash__ | ( | self | ) |
Reimplemented from LayerAttribute.
|
static |
list[PermutationConstraint] get_constraints | ( | self | ) |
def is_complete | ( | self, | |
dict[LayerDim, UnrollFactor] | temporal_loop_sizes | ||
) |
Return wether this temporal ordering matches the given, mandatory loop sizes.
def is_empty | ( | self | ) |
def remove_invalid_layer_dims | ( | self, | |
LayerDimSizes | layer_dim_sizes, | ||
str | layer_name = "" |
||
) |
def to_legacy_format | ( | self | ) |
data |