Apply suggestions from code review

Co-Authored-By: Takenori Yoshimura <[email protected]>

espnet/nets/pytorch_backend/wavenet.py

@@ -249,14 +249,14 @@ def generate(self, x, h, n_samples, interval=None, mode="sampling"):
         This generation based on `Fast WaveNet Generation Algorithm`_.
 
         Args:
-            x (LongTensor): Initial wavenform tensor with the shape  (T).
+            x (LongTensor): Initial waveform tensor with the shape  (T,).
             h (Tensor): Auxiliary feature tensor with the shape  (n_samples + T, n_aux).
             n_samples (int): Number of samples to be generated.
             interval (int, optional): Log interval.
             mode (str, optional): "sampling" or "argmax".
 
         Return:
-            ndarray: Generated quantized wavenform (n_samples).
+            ndarray: Generated quantized waveform (n_samples).
 
         .. _`Fast WaveNet Generation Algorithm`: https://arxiv.org/abs/1611.09482
 

utils/convert_fbank.sh

@@ -21,8 +21,8 @@ Note: <log-dir> defaults to <data-dir>/log, and <fbank-dir> defaults to <data-di
 Options:
   --nj <nj>                  # number of parallel jobs
   --fs <fs>                  # sampling rate
-  --fmax <fmax>              # Maximum frequency
-  --fmin <fmin>              # Minimum frequency
+  --fmax <fmax>              # maximum frequency
+  --fmin <fmin>              # minimum frequency
   --n_fft <n_fft>            # number of FFT points (default=1024)
   --n_shift <n_shift>        # shift size in point (default=256)
   --win_length <win_length>  # window length in point (default=)

utils/generate_wav.sh

@@ -58,7 +58,7 @@ scp=${data}/feats.scp
 
 split_scps=""
 for n in $(seq ${nj}); do
-    split_scps="$split_scps $logdir/feats.$n.scp"
+    split_scps="$split_scps $logdir/feats.${n}.scp"
 done
 
 utils/split_scp.pl ${scp} ${split_scps} || exit 1;
@@ -74,4 +74,4 @@ ${cmd} JOB=1:${nj} ${logdir}/generate_with_wavenet_${name}.JOB.log \
 
 rm ${logdir}/feats.*.scp 2>/dev/null
 
-echo "Succeeded creating wav for $name"
+echo "Succeeded creating wav for ${name}"

utils/generate_wav_from_fbank.py

@@ -78,7 +78,7 @@ def __call__(self, y):
             y (ndarray): Wavnform signal normalized from -1 to 1 (N,).
 
         Returns:
-            y (ndarray): Noise shaped wavnform signal normalized from -1 to 1 (N,).
+            y (ndarray): Noise shaped waveform signal normalized from -1 to 1 (N,).
 
         """
         # check shape and type
@@ -177,7 +177,7 @@ def main():
         h = scaler.transform(lmspc)  # normalize features
 
         # convert to tensor
-        x = torch.tensor(x, dtype=torch.long, device=device)  # (1, )
+        x = torch.tensor(x, dtype=torch.long, device=device)  # (1,)
         h = torch.tensor(h, dtype=torch.float, device=device)  # (T, n_aux)
 
         # get length of waveform

utils/synth_wav.sh

@@ -285,7 +285,7 @@ if [ ${stage} -le 4 ] && [ ${stop_stage} -ge 4 ]; then
     model_corpus=$(echo ${models} | cut -d. -f 1)
     vocoder_model_corpus=$(echo ${vocoder_models} | cut -d. -f 1)
     if [ ${model_corpus} != ${vocoder_model_corpus} ]; then
-        echo "${vocoder_models} does not support ${models} (Due to the sampling rare mismatch)."
+        echo "${vocoder_models} does not support ${models} (Due to the sampling rate mismatch)."
         exit 1
     fi
     download_vocoder_models