From 2f496104ecc7e79627bb17d5d7ac3c2d4cbff539 Mon Sep 17 00:00:00 2001
From: passerbya <hanghang3103@163.com>
Date: Tue, 9 Jul 2024 08:17:34 +0800
Subject: [PATCH] =?UTF-8?q?=E5=8D=8A=E8=A7=92=E5=8F=A5=E5=8F=B7=E4=BC=9A?=
 =?UTF-8?q?=E5=AF=BC=E8=87=B4=E5=90=88=E6=88=90=E5=A4=B1=E8=B4=A5=EF=BC=9A?=
 =?UTF-8?q?RuntimeError:=20torch.cat():=20expected=20a=20non-empty=20list?=
 =?UTF-8?q?=20of=20Tensors?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

text='小明因为感冒,鼻子不通,讲话总带着齉音.'
  File "/usr/local/data/CosyVoice/cosyvoice/cli/cosyvoice.py", line 62, in inference_zero_shot
    return {'tts_speech': torch.concat(tts_speeches, dim=1)}
RuntimeError: torch.cat(): expected a non-empty list of Tensors

原因为self.frontend.text_normalize(tts_text, split=True)返回为空
---
 cosyvoice/utils/frontend_utils.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/cosyvoice/utils/frontend_utils.py b/cosyvoice/utils/frontend_utils.py
index dee829f..bcd637c 100644
--- a/cosyvoice/utils/frontend_utils.py
+++ b/cosyvoice/utils/frontend_utils.py
@@ -74,7 +74,7 @@ def split_paragraph(text: str, tokenize, lang="zh", token_max_n=80, token_min_n=
             return len(tokenize(_text)) < merge_len
 
     if lang == "zh":
-        pounc = ['。', '？', '！', '；', '：', '.', '?', '!', ';']
+        pounc = ['。', '？', '！', '；', '：', '、', '.', '?', '!', ';']
     else:
         pounc = ['.', '?', '!', ';', ':']
     if comma_split: