moses-smt/mosesdecoder
1
1
Fork 0
mirror of https://github.com/moses-smt/mosesdecoder.git synced 2024-09-11 11:25:40 +03:00
Code Issues Packages Projects Releases Wiki Activity

duplication of existing functionality

Browse Source
This commit is contained in:
Rico Sennrich 2015-05-20 17:35:38 +01:00
parent 6aac7ded9a
commit 98ff2382d0
1 changed files with 9 additions and 20 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

29
scripts/ems/experiment.meta
View File

@ -171,13 +171,14 @@ use-parallel-corpus
out: tokenized-corpus
default-name: lm/tok
ignore-unless: parallel-corpus-stem
template: ln -s IN.$output-extension OUT
template: ln -s IN.$output-extension OUT
error: failed to create symbolic link
tokenize
in: raw-corpus
out: tokenized-corpus
default-name: lm/tok
pass-unless: output-tokenizer
ignore-if: parallel-corpus-stem concatenate-files link-file concatenate-files-split link-file-split
ignore-if: parallel-corpus-stem concatenate-files concatenate-files-split
template: $output-tokenizer < IN > OUT
parallelizable: yes
mock-parse
@ -185,14 +186,14 @@ mock-parse
out: mock-parsed-corpus
default-name: lm/mock-parsed
pass-unless: mock-output-parser-lm
ignore-if: concatenate-files link-file concatenate-files-split link-file-split
ignore-if: concatenate-files concatenate-files-split
template: $mock-output-parser-lm < IN > OUT
factorize
in: mock-parsed-corpus
out: factorized-corpus
default-name: lm/factored
pass-unless: factors
ignore-if: concatenate-files link-file concatenate-files-split link-file-split
ignore-if: concatenate-files concatenate-files-split
parallelizable: yes
error: can't open
error: incompatible number of words in factor
@ -201,7 +202,7 @@ lowercase
out: lowercased-corpus
default-name: lm/lowercased
pass-unless: output-lowercaser
ignore-if: output-truecaser concatenate-files link-file concatenate-files-split link-file-split
ignore-if: output-truecaser concatenate-files concatenate-files-split
#only-factor-0: yes
template: $output-lowercaser < IN > OUT
parallelizable: yes
@ -211,7 +212,7 @@ truecase
rerun-on-change: output-truecaser
default-name: lm/truecased
ignore-unless: output-truecaser
ignore-if: concatenate-files link-file concatenate-files-split link-file-split
ignore-if: concatenate-files concatenate-files-split
only-factor-0: yes
template: $output-truecaser -model IN1.$output-extension < IN > OUT
parallelizable: yes
@ -221,14 +222,14 @@ split
rerun-on-change: output-splitter
default-name: lm/split
pass-unless: output-splitter
ignore-if: concatenate-files link-file concatenate-files-split link-file-split
ignore-if: concatenate-files concatenate-files-split
template: $output-splitter -model IN1.$output-extension < IN > OUT
strip
in: split-corpus
out: stripped-corpus
default-name: lm/stripped
pass-unless: mock-output-parser-lm
ignore-if: concatenate-files link-file
ignore-if: concatenate-files
template: $moses-script-dir/training/strip-xml.perl < IN > OUT
concatenate-split
in: concatenate-files-split
@ -242,18 +243,6 @@ concatenate
ignore-unless: concatenate-files
default-name: lm/stripped
template: cat IN > OUT
link-split
in: link-file-split
out: split-corpus
default-name: lm/split
ignore-unless: link-file-split
template: ln -s IN OUT
link
in: link-file
out: stripped-corpus
default-name: lm/stripped
ignore-unless: link-file
template: ln -s IN OUT
train
in: stripped-corpus
out: lm