[TTS]add blank between characters for vits

examples/csmsc/vits/conf/default.yaml

@@ -178,6 +178,8 @@ generator_first: False # whether to start updating generator first
 ##########################################################
 #                OTHER TRAINING SETTING                  #
 ##########################################################
-max_epoch: 1000           # number of epochs
-num_snapshots: 10         # max number of snapshots to keep while training
-seed: 777                 # random seed number
+num_snapshots: 10            # max number of snapshots to keep while training
+train_max_steps: 250000      # Number of training steps. == total_iters / ngpus, total_iters = 1000000
+save_interval_steps: 1000    # Interval steps to save checkpoint.
+eval_interval_steps: 250     # Interval steps to evaluate the network.
+seed: 777                    # random seed number

examples/csmsc/vits/local/preprocess.sh

@@ -4,6 +4,7 @@ stage=0
 stop_stage=100
 
 config_path=$1
+add_blank=$2
 
 if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
     # get durations from MFA's result
@@ -44,6 +45,7 @@ if [ ${stage} -le 3 ] && [ ${stop_stage} -ge 3 ]; then
         --feats-stats=dump/train/feats_stats.npy \
         --phones-dict=dump/phone_id_map.txt \
         --speaker-dict=dump/speaker_id_map.txt \
+        --add-blank=${add_blank} \
         --skip-wav-copy
 
     python3 ${BIN_DIR}/normalize.py \
@@ -52,6 +54,7 @@ if [ ${stage} -le 3 ] && [ ${stop_stage} -ge 3 ]; then
         --feats-stats=dump/train/feats_stats.npy \
         --phones-dict=dump/phone_id_map.txt \
         --speaker-dict=dump/speaker_id_map.txt \
+        --add-blank=${add_blank} \
         --skip-wav-copy
 
     python3 ${BIN_DIR}/normalize.py \
@@ -60,5 +63,6 @@ if [ ${stage} -le 3 ] && [ ${stop_stage} -ge 3 ]; then
         --feats-stats=dump/train/feats_stats.npy \
         --phones-dict=dump/phone_id_map.txt \
         --speaker-dict=dump/speaker_id_map.txt \
+        --add-blank=${add_blank} \
         --skip-wav-copy
 fi

examples/csmsc/vits/local/synthesize_e2e.sh

@@ -3,9 +3,12 @@
 config_path=$1
 train_output_path=$2
 ckpt_name=$3
+add_blank=$4
+
 stage=0
 stop_stage=0
 
+
 if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
     FLAGS_allocator_strategy=naive_best_fit \
     FLAGS_fraction_of_gpu_memory_to_use=0.01 \
@@ -14,5 +17,6 @@ if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
         --ckpt=${train_output_path}/checkpoints/${ckpt_name} \
         --phones_dict=dump/phone_id_map.txt \
         --output_dir=${train_output_path}/test_e2e \
-        --text=${BIN_DIR}/../sentences.txt
+        --text=${BIN_DIR}/../sentences.txt \
+        --add-blank=${add_blank}
 fi

examples/csmsc/vits/run.sh

@@ -10,6 +10,7 @@ stop_stage=100
 conf_path=conf/default.yaml
 train_output_path=exp/default
 ckpt_name=snapshot_iter_153.pdz
+add_blank=true
 
 # with the following command, you can choose the stage range you want to run
 # such as `./run.sh --stage 0 --stop-stage 0`
@@ -18,7 +19,7 @@ source ${MAIN_ROOT}/utils/parse_options.sh || exit 1
 
 if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
     # prepare data
-    ./local/preprocess.sh ${conf_path} || exit -1
+    ./local/preprocess.sh ${conf_path} ${add_blank}|| exit -1
 fi
 
 if [ ${stage} -le 1 ] && [ ${stop_stage} -ge 1 ]; then
@@ -32,5 +33,5 @@ fi
 
 if [ ${stage} -le 3 ] && [ ${stop_stage} -ge 3 ]; then
     # synthesize_e2e, vocoder is pwgan
-    CUDA_VISIBLE_DEVICES=${gpus} ./local/synthesize_e2e.sh ${conf_path} ${train_output_path} ${ckpt_name} || exit -1
+    CUDA_VISIBLE_DEVICES=${gpus} ./local/synthesize_e2e.sh ${conf_path} ${train_output_path} ${ckpt_name} ${add_blank}|| exit -1
 fi

examples/ljspeech/voc0/local/synthesize.sh

@@ -8,5 +8,4 @@ python ${BIN_DIR}/synthesize.py \
     --input=${input_mel_path} \
     --output=${train_output_path}/wavs/ \
     --checkpoint_path=${train_output_path}/checkpoints/${ckpt_name} \
-    --ngpu=1 \
-    --verbose
+    --ngpu=1

paddlespeech/t2s/exps/fastspeech2/normalize.py

@@ -58,30 +58,8 @@ def main():
         "--phones-dict", type=str, default=None, help="phone vocabulary file.")
     parser.add_argument(
         "--speaker-dict", type=str, default=None, help="speaker id map file.")
-    parser.add_argument(
-        "--verbose",
-        type=int,
-        default=1,
-        help="logging level. higher is more logging. (default=1)")
-    args = parser.parse_args()
 
-    # set logger
-    if args.verbose > 1:
-        logging.basicConfig(
-            level=logging.DEBUG,
-            format="%(asctime)s (%(module)s:%(lineno)d) %(levelname)s: %(message)s"
-        )
-    elif args.verbose > 0:
-        logging.basicConfig(
-            level=logging.INFO,
-            format="%(asctime)s (%(module)s:%(lineno)d) %(levelname)s: %(message)s"
-        )
-    else:
-        logging.basicConfig(
-            level=logging.WARN,
-            format="%(asctime)s (%(module)s:%(lineno)d) %(levelname)s: %(message)s"
-        )
-        logging.warning('Skip DEBUG/INFO messages')
+    args = parser.parse_args()
 
     dumpdir = Path(args.dumpdir).expanduser()
     # use absolute path

paddlespeech/t2s/exps/fastspeech2/preprocess.py

@@ -209,11 +209,6 @@ def main():
 
     parser.add_argument("--config", type=str, help="fastspeech2 config file.")
 
-    parser.add_argument(
-        "--verbose",
-        type=int,
-        default=1,
-        help="logging level. higher is more logging. (default=1)")
     parser.add_argument(
         "--num-cpu", type=int, default=1, help="number of process.")
 
@@ -248,10 +243,6 @@ def main():
     with open(args.config, 'rt') as f:
         config = CfgNode(yaml.safe_load(f))
 
-    if args.verbose > 1:
-        print(vars(args))
-        print(config)
-
     sentences, speaker_set = get_phn_dur(dur_file)
 
     merge_silence(sentences)

paddlespeech/t2s/exps/gan_vocoder/normalize.py

@@ -47,30 +47,8 @@ def main():
         default=False,
         action="store_true",
         help="whether to skip the copy of wav files.")
-    parser.add_argument(
-        "--verbose",
-        type=int,
-        default=1,
-        help="logging level. higher is more logging. (default=1)")
-    args = parser.parse_args()
 
-    # set logger
-    if args.verbose > 1:
-        logging.basicConfig(
-            level=logging.DEBUG,
-            format="%(asctime)s (%(module)s:%(lineno)d) %(levelname)s: %(message)s"
-        )
-    elif args.verbose > 0:
-        logging.basicConfig(
-            level=logging.INFO,
-            format="%(asctime)s (%(module)s:%(lineno)d) %(levelname)s: %(message)s"
-        )
-    else:
-        logging.basicConfig(
-            level=logging.WARN,
-            format="%(asctime)s (%(module)s:%(lineno)d) %(levelname)s: %(message)s"
-        )
-        logging.warning('Skip DEBUG/INFO messages')
+    args = parser.parse_args()
 
     dumpdir = Path(args.dumpdir).expanduser()
     # use absolute path

paddlespeech/t2s/exps/gan_vocoder/preprocess.py

@@ -167,11 +167,6 @@ def main():
         required=True,
         help="directory to dump feature files.")
     parser.add_argument("--config", type=str, help="vocoder config file.")
-    parser.add_argument(
-        "--verbose",
-        type=int,
-        default=1,
-        help="logging level. higher is more logging. (default=1)")
     parser.add_argument(
         "--num-cpu", type=int, default=1, help="number of process.")
     parser.add_argument(
@@ -197,10 +192,6 @@ def main():
     with open(args.config, 'rt') as f:
         config = CfgNode(yaml.safe_load(f))
 
-    if args.verbose > 1:
-        print(vars(args))
-        print(config)
-
     sentences, speaker_set = get_phn_dur(dur_file)
     merge_silence(sentences)
 

paddlespeech/t2s/exps/speedyspeech/normalize.py

@@ -50,11 +50,6 @@ def main():
         "--tones-dict", type=str, default=None, help="tone vocabulary file.")
     parser.add_argument(
         "--speaker-dict", type=str, default=None, help="speaker id map file.")
-    parser.add_argument(
-        "--verbose",
-        type=int,
-        default=1,
-        help="logging level. higher is more logging. (default=1)")
 
     parser.add_argument(
         "--use-relative-path",
@@ -63,24 +58,6 @@ def main():
         help="whether use relative path in metadata")
     args = parser.parse_args()
 
-    # set logger
-    if args.verbose > 1:
-        logging.basicConfig(
-            level=logging.DEBUG,
-            format="%(asctime)s (%(module)s:%(lineno)d) %(levelname)s: %(message)s"
-        )
-    elif args.verbose > 0:
-        logging.basicConfig(
-            level=logging.INFO,
-            format="%(asctime)s (%(module)s:%(lineno)d) %(levelname)s: %(message)s"
-        )
-    else:
-        logging.basicConfig(
-            level=logging.WARN,
-            format="%(asctime)s (%(module)s:%(lineno)d) %(levelname)s: %(message)s"
-        )
-        logging.warning('Skip DEBUG/INFO messages')
-
     dumpdir = Path(args.dumpdir).expanduser()
     # use absolute path
     dumpdir = dumpdir.resolve()

paddlespeech/t2s/exps/speedyspeech/preprocess.py

@@ -195,11 +195,6 @@ def main():
 
     parser.add_argument("--config", type=str, help="fastspeech2 config file.")
 
-    parser.add_argument(
-        "--verbose",
-        type=int,
-        default=1,
-        help="logging level. higher is more logging. (default=1)")
     parser.add_argument(
         "--num-cpu", type=int, default=1, help="number of process.")
 
@@ -230,10 +225,6 @@ def main():
     with open(args.config, 'rt') as f:
         config = CfgNode(yaml.safe_load(f))
 
-    if args.verbose > 1:
-        print(vars(args))
-        print(config)
-
     sentences, speaker_set = get_phn_dur(dur_file)
 
     merge_silence(sentences)

paddlespeech/t2s/exps/tacotron2/preprocess.py

@@ -184,11 +184,6 @@ def main():
 
     parser.add_argument("--config", type=str, help="fastspeech2 config file.")
 
-    parser.add_argument(
-        "--verbose",
-        type=int,
-        default=1,
-        help="logging level. higher is more logging. (default=1)")
     parser.add_argument(
         "--num-cpu", type=int, default=1, help="number of process.")
 
@@ -223,10 +218,6 @@ def main():
     with open(args.config, 'rt') as f:
         config = CfgNode(yaml.safe_load(f))
 
-    if args.verbose > 1:
-        print(vars(args))
-        print(config)
-
     sentences, speaker_set = get_phn_dur(dur_file)
 
     merge_silence(sentences)

paddlespeech/t2s/exps/transformer_tts/normalize.py

@@ -51,30 +51,8 @@ def main():
         "--phones-dict", type=str, default=None, help="phone vocabulary file.")
     parser.add_argument(
         "--speaker-dict", type=str, default=None, help="speaker id map file.")
-    parser.add_argument(
-        "--verbose",
-        type=int,
-        default=1,
-        help="logging level. higher is more logging. (default=1)")
-    args = parser.parse_args()
 
-    # set logger
-    if args.verbose > 1:
-        logging.basicConfig(
-            level=logging.DEBUG,
-            format="%(asctime)s (%(module)s:%(lineno)d) %(levelname)s: %(message)s"
-        )
-    elif args.verbose > 0:
-        logging.basicConfig(
-            level=logging.INFO,
-            format="%(asctime)s (%(module)s:%(lineno)d) %(levelname)s: %(message)s"
-        )
-    else:
-        logging.basicConfig(
-            level=logging.WARN,
-            format="%(asctime)s (%(module)s:%(lineno)d) %(levelname)s: %(message)s"
-        )
-        logging.warning('Skip DEBUG/INFO messages')
+    args = parser.parse_args()
 
     # check directory existence
     dumpdir = Path(args.dumpdir).resolve()

paddlespeech/t2s/exps/transformer_tts/preprocess.py

@@ -186,11 +186,6 @@ def main():
         type=str,
         help="yaml format configuration file.")
 
-    parser.add_argument(
-        "--verbose",
-        type=int,
-        default=1,
-        help="logging level. higher is more logging. (default=1)")
     parser.add_argument(
         "--num-cpu", type=int, default=1, help="number of process.")
 
@@ -210,10 +205,6 @@ def main():
         _C = Configuration(_C)
         config = _C.clone()
 
-    if args.verbose > 1:
-        print(vars(args))
-        print(config)
-
     phone_id_map_path = dumpdir / "phone_id_map.txt"
     speaker_id_map_path = dumpdir / "speaker_id_map.txt"