|
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.
|
(for-loop) dimension of a workload layer (e.g. More...


Public Member Functions | |
| def | __init__ (self, str name) |
| "LayerDim" | create_r_version (self) |
Create a new LayerDim instance with is tagged relevant and can be distinguished from non-tagged LayerDims. More... | |
| "LayerDim" | create_ir_version (self) |
Create a new LayerDim instance with is tagged irrelevant and can be distinguished from non-tagged LayerDims. More... | |
Public Member Functions inherited from OperandABC | |
| def | name (self) |
| Protect the class variable from reassignment (as this would invalidate the stored hash value) More... | |
| def | __eq__ (self, "OperandABC" other) |
| def | __hash__ (self) |
| Optimize performance by statically storing the hash. More... | |
| def | __str__ (self) |
| def | __repr__ (self) |
| def | __lt__ (self, "OperandABC" other) |
| def | __ge__ (self, "OperandABC" other) |
| def | __jsonrepr__ (self) |
(for-loop) dimension of a workload layer (e.g.
K, C)
| def __init__ | ( | self, | |
| str | name | ||
| ) |
Reimplemented from OperandABC.
| "LayerDim" create_ir_version | ( | self | ) |
Create a new LayerDim instance with is tagged irrelevant and can be distinguished from non-tagged LayerDims.

| "LayerDim" create_r_version | ( | self | ) |
Create a new LayerDim instance with is tagged relevant and can be distinguished from non-tagged LayerDims.
