add bash scripts

examples/parallelwave_gan/baker/preprocess.py

@@ -202,8 +202,7 @@ def process_sentences(config,
 def main():
     # parse config and args
     parser = argparse.ArgumentParser(
-        description="Preprocess audio and then extract features (See detail in parallel_wavegan/bin/preprocess.py)."
-    )
+        description="Preprocess audio and then extract features .")
     parser.add_argument(
         "--rootdir",
         default=None,

examples/parallelwave_gan/baker/preprocess.sh

@@ -0,0 +1,6 @@
+python preprocess.py --rootdir=~/datasets/BZNSYP/ --dumpdir=dump --num_cpu=20
+python compute_statistics.py --metadata=dump/train/raw/metadata.jsonl --field-name="feats" --dumpdir=dump/train
+
+python normalize.py --metadata=dump/train/raw/metadata.jsonl --dumpdir=dump/train/norm --stats=dump/train/stats.npy
+python normalize.py --metadata=dump/dev/raw/metadata.jsonl --dumpdir=dump/dev/norm --stats=dump/train/stats.npy
+python normalize.py --metadata=dump/test/raw/metadata.jsonl --dumpdir=dump/test/norm --stats=dump/train/stats.npy

examples/parallelwave_gan/baker/run.sh

@@ -0,0 +1,9 @@
+FLAGS_cudnn_exhaustive_search=true \
+FLAGS_conv_workspace_size_limit=4000 \
+CUDA_VISIBLE_DEVICES="3" \
+python train.py \
+    --train-metadata=dump/train/norm/metadata.jsonl \
+    --dev-metadata=dump/dev/norm/metadata.jsonl \
+    --config=conf/debug.yaml \
+    --output-dir=exp/debug \
+    --nprocs=1

examples/speedyspeech/baker/preprocess.py

@@ -221,14 +221,12 @@ def process_sentences(config,
 def main():
     # parse config and args
     parser = argparse.ArgumentParser(
-        description="Preprocess audio and then extract features (See detail in parallel_wavegan/bin/preprocess.py)."
-    )
+        description="Preprocess audio and then extract features.")
     parser.add_argument(
         "--rootdir",
         default=None,
         type=str,
-        help="directory including wav files. you need to specify either scp or rootdir."
-    )
+        help="directory to baker dataset.")
     parser.add_argument(
         "--dumpdir",
         type=str,

examples/speedyspeech/baker/preprocess.sh

@@ -0,0 +1,6 @@
+python preprocess.py --rootdir=~/datasets/BZNSYP/ --dumpdir=dump --num_cpu=20
+python compute_statistics.py --metadata=dump/train/raw/metadata.jsonl --field-name="feats" --dumpdir=dump/train
+
+python normalize.py --metadata=dump/train/raw/metadata.jsonl --dumpdir=dump/train/norm --stats=dump/train/stats.npy
+python normalize.py --metadata=dump/dev/raw/metadata.jsonl --dumpdir=dump/dev/norm --stats=dump/train/stats.npy
+python normalize.py --metadata=dump/test/raw/metadata.jsonl --dumpdir=dump/test/norm --stats=dump/train/stats.npy

examples/speedyspeech/baker/run.sh

@@ -0,0 +1,7 @@
+CUDA_VISIBLE_DEVICES="2" \
+python train.py \
+    --train-metadata=dump/train/norm/metadata.jsonl \
+    --dev-metadata=dump/dev/norm/metadata.jsonl \
+    --config=conf/default.yaml \
+    --output-dir=exp/default \
+    --nprocs=1

examples/speedyspeech/baker/synthesize.sh

@@ -0,0 +1,10 @@
+python synthesize.py \
+  --speedyspeech-config=conf/default.yaml \
+  --speedyspeech-checkpoint=exp/debug/checkpoints/snapshot_iter_91800.pdz \
+  --speedyspeech-stat=dump/train/stats.npy \
+  --pwg-config=../../parallelwave_gan/baker/conf/default.yaml \
+  --pwg-params=../../parallelwave_gan/baker/converted.pdparams \
+  --pwg-stat=../../parallelwave_gan/baker/dump/train/stats.npy \
+  --test-metadata=dump/test/norm/metadata.jsonl \
+  --output-dir=exp/debug/test \
+  --device="gpu"

examples/speedyspeech/baker/synthesize_e2e.sh

@@ -0,0 +1,10 @@
+python synthesize_e2e.py \
+  --speedyspeech-config=conf/default.yaml \
+  --speedyspeech-checkpoint=exp/debug/checkpoints/snapshot_iter_91800.pdz \
+  --speedyspeech-stat=dump/train/stats.npy \
+  --pwg-config=../../parallelwave_gan/baker/conf/default.yaml \
+  --pwg-params=../../parallelwave_gan/baker/converted.pdparams \
+  --pwg-stat=../../parallelwave_gan/baker/dump/train/stats.npy \
+  --text=sentences.txt \
+  --output-dir=exp/e2e \
+  --device="gpu"