persistent-function-cache 0.2.0

Persistent cache for expensive functions

Cache

Usage

Use

from persistent_cache import cache

@cache
def expensive_function(..):
    ..

to cache the result of a function

The cache key for the result is determined by:

• the function signature
• the implementation of the function
• the values of the function arguments
  • custom transformations/reductions can be specified

Advantages compared to existing solutions:

• the cache in invalidated when the behavior of the function changes
• Each cache value is saved to a separate location. Only values that are effectively needed are loaded.
• works with function arguments of any complex data type.
• configurable: custom transformations/reductions can be specified based on the object type.
• 3 custom transformation groups available out-of-the-box:
  • from persistent_cache import cache
  • from persistent_cache.caches.deep_learning import cache
  • from persistent_cache.caches.speedup_deep_learning import cache`

Installation

pip install persistent-function-cache