[
https://issues.apache.org/jira/browse/OPENNLP-569?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Vinh Khuc updated OPENNLP-569:
------------------------------
Attachment: qn-trainer.params
nl-per-train.log
nl-per-testb.log
nl-per-testa.log
nl-per.train
nl-per.testb
nl-per.testa
Added training set, test sets from
https://issues.apache.org/jira/browse/OPENNLP-338 and the output logs of
QNTrainer after using this patch.
Assuming the current directory is trunk/opennlp-tools, the following commands
were used to create the logs
bin\opennlp TokenNameFinderTrainer -lang nl -data nl-per.train -params
qn-trainer.params -model qn-test.bin
bin\opennlp TokenNameFinderEvaluator -model qn-test.bin -data nl-per.testa
bin\opennlp TokenNameFinderEvaluator -model qn-test.bin -data nl-per.testb
> Fix remaining issue in L-BFGS parameter estimation to get it stable
> -------------------------------------------------------------------
>
> Key: OPENNLP-569
> URL: https://issues.apache.org/jira/browse/OPENNLP-569
> Project: OpenNLP
> Issue Type: Improvement
> Components: Machine Learning
> Affects Versions: tools-1.5.3, maxent-3.0.3
> Reporter: Joern Kottmann
> Assignee: Hyosup Shim
> Attachments: LBFGS.patch, nl-per-testa.log, nl-per-testb.log,
> nl-per-train.log, nl-per.testa, nl-per.testb, nl-per.train, qn-trainer.params
>
>
> Enhance the L-BFGS parameter estimation code to at least perform as well as
> the GIS training. The work on this issue should bring the implementation to a
> level where the experimental flag can be removed.
> For remaining problems see this issue: OPENNLP-338.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
