diff options
| author | trevor.cohn <trevor.cohn@ec762483-ff6d-05da-a07a-a48fb63a330f> | 2010-06-28 19:34:58 +0000 |
|---|---|---|
| committer | trevor.cohn <trevor.cohn@ec762483-ff6d-05da-a07a-a48fb63a330f> | 2010-06-28 19:34:58 +0000 |
| commit | 89afb013e40bf9c749956420e77ac72773874844 (patch) | |
| tree | 2e18cf00d768507ae5cbc4cb085e22165f3b6fe6 /gi/posterior-regularisation/README | |
| parent | c1fd8f9ba0c8a6f5ab07a3390d3f1a3910a3267e (diff) | |
First bits of code for PR training
git-svn-id: https://ws10smt.googlecode.com/svn/trunk@44 ec762483-ff6d-05da-a07a-a48fb63a330f
Diffstat (limited to 'gi/posterior-regularisation/README')
| -rw-r--r-- | gi/posterior-regularisation/README | 3 |
1 files changed, 3 insertions, 0 deletions
diff --git a/gi/posterior-regularisation/README b/gi/posterior-regularisation/README new file mode 100644 index 00000000..a3d54ffc --- /dev/null +++ b/gi/posterior-regularisation/README @@ -0,0 +1,3 @@ + 557 ./cdec_extools/extractor -i btec/split.zh-en.al -c 500000 -L 12 -C | sort -t $'\t' -k 1 | ./cdec_extools/mr_stripe_rule_reduce > btec.concordance + 559 wc -l btec.concordance + 588 cat btec.concordance | sed 's/.* //' | awk '{ for (i=1; i < NF; i++) { x=substr($i, 1, 2); if (x == "C=") printf "\n"; else if (x != "||") printf "%s ", $i; }; printf "\n"; }' | sort | uniq | wc -l |