From ad3391bd96b9a40bd17bf26c93daf47b9bbb5849 Mon Sep 17 00:00:00 2001
From: James Betker <jbetker@gmail.com>
Date: Sat, 14 Aug 2021 20:42:01 -0600
Subject: [PATCH] Fix nan issue when interpolating audio

---
 codes/data/audio/nv_tacotron_dataset.py | 14 ++++++++------
 codes/models/tacotron2/text/__init__.py |  2 +-
 2 files changed, 9 insertions(+), 7 deletions(-)

diff --git a/codes/data/audio/nv_tacotron_dataset.py b/codes/data/audio/nv_tacotron_dataset.py
index 8a9d6361..98f085af 100644
--- a/codes/data/audio/nv_tacotron_dataset.py
+++ b/codes/data/audio/nv_tacotron_dataset.py
@@ -79,20 +79,22 @@ class TextMelLoader(torch.utils.data.Dataset):
         if not self.load_mel_from_disk:
             if filename.endswith('.wav'):
                 audio, sampling_rate = load_wav_to_torch(filename)
-                audio = (audio / self.max_wav_value).clip(-1,1)
+                audio = (audio / self.max_wav_value)
             else:
                 audio, sampling_rate = audio2numpy.audio_from_file(filename)
                 audio = torch.tensor(audio)
-                audio = (audio.squeeze().clip(-1,1))
 
             if sampling_rate != self.input_sample_rate:
                 if sampling_rate < self.input_sample_rate:
                     print(f'{filename} has a sample rate of {sampling_rate} which is lower than the requested sample rate of {self.input_sample_rate}. This is not a good idea.')
-                audio = torch.nn.functional.interpolate(audio.unsqueeze(0).unsqueeze(1), scale_factor=self.input_sample_rate/sampling_rate, mode='area', recompute_scale_factor=False).squeeze()
-            if (audio.min() < -1).any() or (audio.max() > 1).any():
-                print(f"Error with audio ranging for {filename}; min={audio.min()} max={audio.max()}")
+                audio_norm = torch.nn.functional.interpolate(audio.unsqueeze(0).unsqueeze(1), scale_factor=self.input_sample_rate/sampling_rate, mode='nearest', recompute_scale_factor=False).squeeze()
+            else:
+                audio_norm = audio
+            if audio_norm.std() > 1:
+                print(f"Something is very wrong with the given audio. std_dev={audio_norm.std()}. file={filename}")
                 return None
-            audio_norm = audio.unsqueeze(0)
+            audio_norm.clip_(-1, 1)
+            audio_norm = audio_norm.unsqueeze(0)
             audio_norm = torch.autograd.Variable(audio_norm, requires_grad=False)
             if self.input_sample_rate != self.sampling_rate:
                 ratio = self.sampling_rate / self.input_sample_rate
diff --git a/codes/models/tacotron2/text/__init__.py b/codes/models/tacotron2/text/__init__.py
index a392b10a..f8a3228f 100644
--- a/codes/models/tacotron2/text/__init__.py
+++ b/codes/models/tacotron2/text/__init__.py
@@ -76,4 +76,4 @@ def _arpabet_to_sequence(text):
 
 
 def _should_keep_symbol(s):
-  return s in _symbol_to_id and s is not '_' and s is not '~'
+  return s in _symbol_to_id and s != '_' and s != '~'