Hi, The error happens during word alignment, so it is independent of using a phrase-based or hierarchical or syntax model.
Can you check if the GIZA++ binary could be found by the script? -phi 2010/6/24 liukai <[email protected]>: > Hi Everyone: > > As the manual : "In short, training uses the identical training script as > phrase-based models. When running train-model.perl, you will have to specify > additional parameters -hierarchical -glue-grammar. You typically will also > reduce the number of lexical items in the grammar with -max-phrase-length 5" > > And I do so (the training corpus has been transform to <tree > label="XX">...... like): > > trainPrefix="train/fbis" > trainF="treecn" > trainE="en" > ${mosesScript}/training/train-factored-phrase-model.perl \ > --scripts-root-dir "${mosesScript}" \ > --root-dir "${trainDir}" \ > --corpus "${dataDir}/${trainPrefix}" \ > --f ${trainF} \ > --e ${trainE} \ > --max-phrase-length 5 \ > --reordering msd-bidirectional-fe \ > --alignment grow-diag-final-and \ > --lm 0:5:"${lm}":0 \ > --first-step 1 \ > --last-step 9 \ > --parallel \ > -source-syntax \ > -hierarchical \ > -glue-grammar > > And it has some errors , last one is : > ERROR: Giza did not produce the output file /home3/mht/liukai/moses-test//kc-train-final/giza.en-treecn/en-treecn.A3.final. Is your corpus clean (reasonably-sized sentences)? at /home3/mht/liukai/moses-decoder/trunk//scripts/target/scripts-20100622-1030//training/train-factored-phrase-model.perl line 961. > > So I did not get the phrase table . And how can I get it ? > > > 2010-06-24 > ________________________________ > liukai > _______________________________________________ > Moses-support mailing list > [email protected] > http://mailman.mit.edu/mailman/listinfo/moses-support > > _______________________________________________ Moses-support mailing list [email protected] http://mailman.mit.edu/mailman/listinfo/moses-support
