added import for italian M-AILABS dataset (#21)

MozillaItalia · Oct 10, 2019 · 76e1546 · 76e1546
1 parent 2332e03
commit 76e1546
Show file tree

Hide file tree

Showing 2 changed files with 18 additions and 0 deletions.
diff --git a/DeepSpeech/import_m-ailabs.sh b/DeepSpeech/import_m-ailabs.sh
@@ -0,0 +1,16 @@
+#!/usr/bin/env bash
+set -xe
+
+pushd $HOME/ds/
+	if [ "${ENGLISH_COMPATIBLE}" = "1" ]; then
+		IMPORT_AS_ENGLISH="--normalize"
+	fi;
+
+	if [ ! -f "/mnt/extracted/data/M-AILABS/it_IT/it_IT_train.csv" ]; then
+		# added --skiplist nothing to avoid issues with the importer TODO: find a proper solution
+		python bin/import_m-ailabs.py ${IMPORT_AS_ENGLISH} \
+		    --skiplist nothing\
+			--language it_IT                                \
+			/mnt/extracted/data/M-AILABS/
+	fi;
+popd
diff --git a/DeepSpeech/run_it.sh b/DeepSpeech/run_it.sh
@@ -4,6 +4,8 @@ set -xe
 
 $HOME/import_cvit.sh
 
+$HOME/import_m-ailabs.sh
+
 $HOME/generate_alphabet.sh
 
 $HOME/build_lm.sh