From e451eb799a5bccd53dfd4b86cf66a4668b0088b7 Mon Sep 17 00:00:00 2001
From: shixian.shi <shixian.shi@alibaba-inc.com>
Date: 星期三, 06 三月 2024 15:31:47 +0800
Subject: [PATCH] infer for word punc model

---
 funasr/datasets/audio_datasets/preprocessor.py |    9 +++++----
 1 files changed, 5 insertions(+), 4 deletions(-)

diff --git a/funasr/datasets/audio_datasets/preprocessor.py b/funasr/datasets/audio_datasets/preprocessor.py
index a3ba3a5..ab75140 100644
--- a/funasr/datasets/audio_datasets/preprocessor.py
+++ b/funasr/datasets/audio_datasets/preprocessor.py
@@ -26,10 +26,11 @@
 			return waveform
 		speed = random.choice(self.speed_perturb)
 		if speed != 1.0:
-			with torch.no_grad():
-				waveform, _ = torchaudio.sox_effects.apply_effects_tensor(
-					torch.tensor(waveform).view(1, -1), fs, [['speed', str(speed)], ['rate', str(fs)]])
-				waveform = waveform.view(-1)
+			if not isinstance(waveform, torch.Tensor):
+				waveform = torch.tensor(waveform)
+			waveform, _ = torchaudio.sox_effects.apply_effects_tensor(
+				waveform.view(1, -1), fs, [['speed', str(speed)], ['rate', str(fs)]])
+			waveform = waveform.view(-1)
 			
 		return waveform
 

--
Gitblit v1.9.1