From 5e63baf984e7de64952aa6d5d396698979397d76 Mon Sep 17 00:00:00 2001
From: Tilman Kamp <5991088+tilmankamp@users.noreply.github.com>
Date: Fri, 24 Apr 2020 08:43:55 +0200
Subject: [PATCH] Fix for gaps augmentation

---
 training/deepspeech_training/util/signal_augmentations.py | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/training/deepspeech_training/util/signal_augmentations.py b/training/deepspeech_training/util/signal_augmentations.py
index f9156c39ac..d69622d435 100644
--- a/training/deepspeech_training/util/signal_augmentations.py
+++ b/training/deepspeech_training/util/signal_augmentations.py
@@ -141,7 +141,8 @@ def apply(self, sample, clock):
         n_gaps = pick_value_from_range(self.n_gaps, clock=clock)
         for _ in range(n_gaps):
             size = pick_value_from_range(self.size, clock=clock)
-            offset = max(0, random.randint(0, len(audio) - size - 1))
+            size = min(size, len(audio) // 10)  # a gap should never exceed 10 percent of the audio
+            offset = random.randint(0, max(0, len(audio) - size - 1))
             audio[offset:offset + size] = 0
         sample.audio = audio