summaryrefslogtreecommitdiff
path: root/dtrain/job2/hadoop-streaming-job.sh
diff options
context:
space:
mode:
Diffstat (limited to 'dtrain/job2/hadoop-streaming-job.sh')
-rwxr-xr-xdtrain/job2/hadoop-streaming-job.sh23
1 files changed, 23 insertions, 0 deletions
diff --git a/dtrain/job2/hadoop-streaming-job.sh b/dtrain/job2/hadoop-streaming-job.sh
new file mode 100755
index 00000000..9ee70a33
--- /dev/null
+++ b/dtrain/job2/hadoop-streaming-job.sh
@@ -0,0 +1,23 @@
+#!/bin/bash
+
+HADOOP_HOME=/usr/lib/hadoop-0.20
+JAR=contrib/streaming/hadoop-streaming-0.20.2-cdh3u1.jar
+HSTREAMING="$HADOOP_HOME/bin/hadoop jar $HADOOP_HOME/$JAR"
+
+IN=in/nc-wmt11-de-en-dyer-cs-joshua.tok.lc.fixamp1.loo.psg.dtrain
+OUT=out/nc-wmt11-de-en-dyer-cs-joshua.tok.lc.fixamp1.loo.psg.dtrain-weights.50
+
+$HSTREAMING \
+ -mapper "dtrain.sh" \
+ -reducer "avgweights.rb" \
+ -input $IN \
+ -output $OUT \
+ -file avgweights.rb \
+ -file dtrain.sh \
+ -file dtrain \
+ -file dtrain.ini \
+ -file cdec.ini \
+ -file nc-wmt11.en.srilm.3.gz \
+ -jobconf mapred.reduce.tasks=1 \
+ -jobconf mapred.max.map.failures.percent=100
+