nested_compression
Bases: nested_expansion
The nested data compression function.
It performs the data compression of multiple compression functions, and conctatnates their compressions to define wider compressions of the input data vector. This class inherits from the nested_expansion class.
...
Notes
Formally, given the \(n\) different compression functions (1) \(\kappa_1: {R}^{m} \to {R}^{d_1}\), (2) \(\kappa_2: {R}^{d_1} \to {R}^{d_2}\), \(\cdots\), (n) \(\kappa_n: {R}^{d_{n-1}} \to {R}^{D}\), we can represent their nested data compression function \(\kappa: {R}^{m} \to {R}^D\) as follows: $$ \begin{equation} \kappa(\mathbf{x}) = \kappa_{n} \left( \kappa_{n-1} \left( \cdots \kappa_2 \left( \kappa_{1} \left( \mathbf{x} \right) \right) \right) \right) \in {R}^{D}. \end{equation} $$ where the compression output dimension \(D\) is determined by the last compression function \(\kappa_n\).
Attributes:
Name | Type | Description |
---|---|---|
name |
(str, optional)
|
The name of the nested compression function. Defaults to 'nested_compression'. |
Methods:
Name | Description |
---|---|
__init__ |
It performs the initialization of the compression function. |
Source code in tinybig/compression/nested_compression.py
__init__(name='nested_compression', *args, **kwargs)
The initialization method of the nested compression function.
It initializes an nested compression object based on the input function name. This method will also call the initialization method of the base class as well.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
name
|
str
|
The name of the nested compression function. Defaults to 'nested_compression'. |
'nested_compression'
|
Returns:
Type | Description |
---|---|
transformation
|
The nested data compression function. |