Support lazy insertions and retractions #58

Closed
rbrush opened this Issue Jun 16, 2014 · 1 comment

Comments

Projects
None yet
1 participant
@rbrush
Collaborator

rbrush commented Jun 16, 2014

Rules that have many activations, inserting many derived facts into the working memory, could be made more efficient. Today we process each activation independently, but lazily firing them as batches after a set of activations would allow them to take advantage of the batch network optimizations more efficiently.

@rbrush rbrush added this to the 0.9.0 milestone Jan 8, 2015

rbrush added a commit that referenced this issue Jan 9, 2015

@rbrush rbrush modified the milestones: 0.9.0, 0.8.1 Jan 9, 2015

@rbrush

This comment has been minimized.

Show comment
Hide comment
@rbrush

rbrush Jan 9, 2015

Collaborator

The above commits improve performance based on some real-world use cases as expected. Closing this issue.

Collaborator

rbrush commented Jan 9, 2015

The above commits improve performance based on some real-world use cases as expected. Closing this issue.

@rbrush rbrush closed this Jan 9, 2015

@rbrush rbrush modified the milestones: 0.8.1, 0.8.2 Jan 9, 2015

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment