Hi Ondrej, I have a phrase table with 10 features and 2 LMs and I'm trying to tune this system with the new mert-moses.pl, which it looks like you've been hacking on recently. There appear to be sanity checks in about 3 different places in the script now (which is nice). However, when I specify an appropriate --lambdas string (see below), the script complains that $used_triples doesn't have the expected number of weights. However, it looks like these sanity checks are invoked before information from --lambdas has been added into $used_triples. For now, the system seems to be tuning as expected (all features are being tuned) by simply commenting these checks out. Any thoughts?
Cheers, Jon --lambdas="d:0.2,-3.0-3.0;0.2,-3.0-3.0;0.2,-3.0-3.0;0.2,-3.0-3.0;0.2,-3.0-3.0;0.2,-3.0-3.0;0.2,-3.0-3.0 lm:1,-3.0-3.0;1,-3.0-3.0 tm:0.3,-3.0-3.0;0.3,-3.0-3.0;0.3,-3.0-3.0;0.3,-3.0-3.0;0.3,-3.0-3.0;0.3,-3.0-3.0;0.3,-3.0-3.0;0.3,-3.0-3.0;0.3,-3.0-3.0;0.3,-3.0-3.0; w:0.1,-2.0-2.0" _______________________________________________ Moses-support mailing list Moses-support@mit.edu http://mailman.mit.edu/mailman/listinfo/moses-support
