Age | Commit message (Collapse) | Author |
|
|
|
|
|
|
|
still not threadsafe
|
|
|
|
|
|
|
|
|
|
|
|
|
|
from the training directory to the root directory
|
|
|
|
Copies the run*.moses.ini with the highest development BLEU to
moses.ini, instead of copying the weights from the last run of
optimization.
Recommended when using kbmira (and probably pro).
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
- Add the descriptions to the usage.
- Initialize the variable $sum.
- Close file handle.
|
|
|
|
|
|
|
|
Convert relative to absoluate paths using File::Spec::Unix
|
|
|
|
|
|
|
|
|
|
|
|
|
|
-script-root-dir
|
|
|
|
|
|
stop namespace pollution
|
|
|
|
|
|
|
|
Regression tests all check out, and kbmira seems to work fine
on a Hansard French->English task.
HypPackEnumerator class may be of interest to pro.cpp and future
optimizers, as it abstracts a lot of the boilerplate involved in
enumerating multiple k-best lists.
MiraWeightVector is not really mira-specific - just a weight vector
that enables efficient averaging. Could be useful to a perceptron
as well. Same goes for MiraFeatureVector.
Interaction with sparse features is written, but untested.
|