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.
|
Class to validate user-given mappings from yaml file. More...
Public Member Functions | |
def | __init__ (self, Any data) |
Initialize Validator object, assign schema and store normalize user-given data. More... | |
def | normalized_data (self) |
Return normalized, user-provided data. More... | |
def | invalidate (self, str extra_msg) |
bool | validate (self) |
Validate the user-provided accelerator data. More... | |
None | validate_single_mapping (self, dict[str, Any] layer_data) |
Public Attributes | |
validator | |
schema | |
is_valid | |
Static Public Attributes | |
dictionary | SCHEMA_SINGLE |
Class to validate user-given mappings from yaml file.
def __init__ | ( | self, | |
Any | data | ||
) |
Initialize Validator object, assign schema and store normalize user-given data.
def invalidate | ( | self, | |
str | extra_msg | ||
) |
def normalized_data | ( | self | ) |
Return normalized, user-provided data.
bool validate | ( | self | ) |
Validate the user-provided accelerator data.
Log a critical warning when invalid data is encountered and return true iff valid.
None validate_single_mapping | ( | self, | |
dict[str, Any] | layer_data | ||
) |
is_valid |
schema |
|
static |
validator |