openhub.net
Black Duck Software, Inc.
Open Hub
Follow @
OH
Sign In
Join Now
Projects
People
Organizations
Tools
Blog
BDSA
Projects
People
Projects
Organizations
P
pystruct
Settings
|
Report Duplicate
0
I Use This!
×
Login Required
Log in to Open Hub
Remember Me
Inactive
Commits
: Listings
Analyzed
about 18 hours
ago. based on code collected
about 18 hours
ago.
Feb 15, 2025 — Feb 15, 2026
Showing page 5 of 43
Search / Filter on:
Commit Message
Contributor
Files Modified
Lines Added
Lines Removed
Code Location
Date
ok
Jean-Luc Meunier
More...
about 9 years ago
ad3 can now deal with multitype CRFs
meunier
More...
about 9 years ago
- the 2 inference methods are now inherited!! - loss-augmentation method
meunier
More...
about 9 years ago
- inherits now from CRF - setInferenceMethod method available
meunier
More...
about 9 years ago
the loss augmentation of the unaires is noz encapsulated in a method for possible specialization by sub class
Jean-Luc Meunier
More...
about 9 years ago
0.3.2
Jean-Luc Meunier
More...
about 9 years ago
code cleaning
Jean-Luc Meunier
More...
about 9 years ago
passed
Jean-Luc Meunier
More...
about 9 years ago
same as before
Jean-Luc Meunier
More...
about 9 years ago
- code cleaning - convenience method flattenY unflattenY - continuous_loss re-implemented for the case of the relaxed marginals returned as lists
Jean-Luc Meunier
More...
about 9 years ago
- supports the case of an inference method returning a tuple of LIST OF marginals, in relaxed mode
Jean-Luc Meunier
More...
about 9 years ago
ad3+ inference returns a list of unaries and a list of pairwise (smaller total size than all this in a big matrix with lots of zeros)
Jean-Luc Meunier
More...
about 9 years ago
- test ok - pairwise is a list
Jean-Luc Meunier
More...
about 9 years ago
the pairwise is now a list per type
Jean-Luc Meunier
More...
about 9 years ago
- ad3+ is the only inference method - a list of unary per type
Jean-Luc Meunier
More...
about 9 years ago
in multitype mode, we pass a list of unaries instead of a big matrix for all types (with lots of 0s)
Jean-Luc Meunier
More...
about 9 years ago
v=0.3.1 !!
Jean-Luc Meunier
More...
about 9 years ago
little fix v=0.3.1
Jean-Luc Meunier
More...
about 9 years ago
- all inference methods are tolarated
Jean-Luc Meunier
More...
about 9 years ago
- the 2 methods for inference are now in this module (TypedCRF is abstract class) - we tolerate all inference method at training time. Not sure it helps in any thing, actually.
Jean-Luc Meunier
More...
about 9 years ago
- ad3+ is the preferred and by default inference method - inference and loss_augmented_inference method moved to NodeTYpeEdgeFeatureCRF
Jean-Luc Meunier
More...
about 9 years ago
fixed bug due to comment...
Jean-Luc Meunier
More...
about 9 years ago
v0.3.0 requiring ad3 2.1.0
Jean-Luc Meunier
More...
about 9 years ago
what's new in 0.3.0
Jean-Luc Meunier
More...
about 9 years ago
change version number to 0.3.0 added me as author
Jean-Luc Meunier
More...
about 9 years ago
declare TypedCRF
Jean-Luc Meunier
More...
about 9 years ago
fixed method doc
Jean-Luc Meunier
More...
about 9 years ago
Class to suport multi-type CRF graphs
Jean-Luc Meunier
More...
about 9 years ago
new inference_ad3plus method InferenceException exception for the new method
Jean-Luc Meunier
More...
about 9 years ago
added inference_ad3plus method and InferenceException exception
Jean-Luc Meunier
More...
about 9 years ago
←
1
2
3
4
5
6
7
8
9
…
42
43
→
This site uses cookies to give you the best possible experience. By using the site, you consent to our use of cookies. For more information, please see our
Privacy Policy
Agree