Running multiple Nengo models using python multiprocessing

General Discussion
#1

I am running multiple nengo models using python multiprocessing library and I got following user warning

nengo/cache.py:447: UserWarning: Decoder cache index could not acquire lock. Cache index was not synced.

Could will it cause any complications sooner or later?

#2

No, that should be fine (it just means that you won’t benefit from the caching). But you could add nengo.rc.set("decoder_cache", "enabled", False) if you want to explicitly disable the cache and avoid those warnings.

1 Like
#3

Found this thread while googling for documentation. You’ll want to use:

nengo.rc.set("decoder_cache", "enabled", "False")

("False" instead of False) otherwise you’ll get:

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-2-4005b404c25d> in <module>
      1 import nengo
----> 2 nengo.rc.set("decoder_cache", "enabled", False)

~/anaconda3/envs/py36/lib/python3.6/configparser.py in set(self, section, option, value)
   1190         """Set an option.  Extends RawConfigParser.set by validating type and
   1191         interpolation syntax on the value."""
-> 1192         self._validate_value_types(option=option, value=value)
   1193         super().set(section, option, value)
   1194 

~/anaconda3/envs/py36/lib/python3.6/configparser.py in _validate_value_types(self, section, option, value)
   1175         if not self._allow_no_value or value:
   1176             if not isinstance(value, str):
-> 1177                 raise TypeError("option values must be strings")
   1178 
   1179     @property

TypeError: option values must be strings