Merge pull request #147 from lissyx/ds-0.9

Ds 0.9
common-voice · Dec 3, 2020 · 5a0f61b · 5a0f61b
2 parents 4be6d3e + e9b5a96
commit 5a0f61b
Show file tree

Hide file tree

Showing 6 changed files with 61 additions and 6 deletions.
diff --git a/DeepSpeech/Dockerfile.train b/DeepSpeech/Dockerfile.train
@@ -1,8 +1,8 @@
 FROM nvidia/cuda:10.0-cudnn7-runtime-ubuntu18.04
 
 ARG ds_repo=mozilla/DeepSpeech
-ARG ds_branch=3ebcbdea91f1d60fc9e92c35da7b040f312246df
-ARG ds_sha1=3ebcbdea91f1d60fc9e92c35da7b040f312246df
+ARG ds_branch=4270e22fe02f4fa7430a721ac917f6353c36f455
+ARG ds_sha1=4270e22fe02f4fa7430a721ac917f6353c36f455
 ARG cc_repo=mozilla/CorporaCreator
 ARG cc_sha1=73622cf8399f8e634aee2f0e76dacc879226e3ac
 ARG kenlm_repo=kpu/kenlm
@@ -90,6 +90,7 @@ RUN apt-get -qq update && apt-get -qq install -y --no-install-recommends \
     curl \
     wget \
     git \
+    ffmpeg \
     python3 \
     python3-pip \
     ca-certificates \
@@ -170,14 +171,16 @@ COPY --chown=trainer:trainer corpora.patch $CC_DIR
 
 RUN patch -p1 < corpora.patch
 
-# Avoid "error: pandas 1.1.0 is installed but pandas==1.0.5 is required by {'modin'}"
-RUN pip install pandas==1.0.5
-
 # error: parso 0.8.0 is installed but parso<0.8.0,>=0.7.0 is required by {'jedi'}
 RUN pip install parso==0.7.0
 
+RUN pip install modin[all]
+
 RUN python setup.py install
 
+# For CC PMF importer
+RUN pip install num2words
+
 WORKDIR $HOMEDIR
 
 ENV PATH="$HOMEDIR/kenlm/build/bin/:$PATH"

diff --git a/DeepSpeech/checks.sh b/DeepSpeech/checks.sh
@@ -27,6 +27,8 @@ done;
 
 mkdir /mnt/extracted/data/ || true
 
+python -c "import tensorflow as tf; tf.test.is_gpu_available()"
+
 # Checking with basic LDC93S1 before running into heavy-load
 pushd $HOME/ds/
 	./bin/run-tc-ldc93s1_new.sh 2 16000

diff --git a/DeepSpeech/fr/import_ccpmf.sh b/DeepSpeech/fr/import_ccpmf.sh
@@ -0,0 +1,14 @@
+#!/bin/bash
+
+set -xe
+
+pushd $HOME/ds/
+	if [ ! -f "/mnt/extracted/data/ccpmf/transcriptionsXML_audioMP3_MEFR_CCPMF_2012-2020/ccpmf_train.csv" ]; then
+		#  Hot patching like that.
+		sed -ri 's/MAX_SECS = .*/MAX_SECS = 4.5/g' bin/import_ccpmf.py
+
+		python bin/import_ccpmf.py \
+			${IMPORTERS_VALIDATE_LOCALE} \
+			/mnt/extracted/data/ccpmf/
+	fi;
+popd
diff --git a/DeepSpeech/fr/importers.sh b/DeepSpeech/fr/importers.sh
@@ -11,3 +11,5 @@ import_trainingspeech.sh
 import_slr57.sh
 
 ../import_m-ailabs.sh
+
+import_ccpmf.sh
diff --git a/DeepSpeech/fr/metadata.sh b/DeepSpeech/fr/metadata.sh
@@ -4,7 +4,7 @@ set -xe
 
 export METADATA_AUTHOR="DeepSpeech-FR-Team"
 export METADATA_MODEL_NAME="deepspeech-fr"
-export METADATA_MODEL_VERSION="0.5"
+export METADATA_MODEL_VERSION="0.6"
 export METADATA_CONTACT_INFO="https://discourse.mozilla.org/c/voice/fr"
 export METADATA_LICENSE="MIT-0"
 export METADATA_LANGUAGE="fr-FR"

diff --git a/DeepSpeech/fr/validate_label.py b/DeepSpeech/fr/validate_label.py
@@ -11,6 +11,34 @@ def validate_label(label):
         return None
 
     skip_foreign_chars = [
+        'ʔ',
+        'ε',
+        'ι',
+        'ο',
+        'ό',
+        'ρ',
+        'ς',
+        'ψ',
+        'գ',
+        'զ',
+        'ا',
+        'ب',
+        'د',
+        'ر',
+        'ل',
+        'ن',
+        'و',
+        'ي',
+        'ቀ',
+        'ወ',
+        'う',
+        'ゔ',
+        'へ',
+        'ま',
+        'め',
+        'や',
+        '貴',
+        '青',
         'い',
         'た',
         'つ',
@@ -25,6 +53,12 @@ def validate_label(label):
         '杜',
         '美',
         '馆',
+        '삼',
+        '고',
+        '생',
+        '기',
+        '집',
+        '먹',
     ]
 
     for skip in skip_foreign_chars: