From 476dc3f30c014e0d2ebdc46ce0283ddbfe63eeb8 Mon Sep 17 00:00:00 2001
From: VirtuosoQ <2416050435@qq.com>
Date: 星期日, 28 四月 2024 16:37:54 +0800
Subject: [PATCH] 16:37 java_http_client
---
funasr/utils/load_utils.py | 22 +++++++++++++++-------
1 files changed, 15 insertions(+), 7 deletions(-)
diff --git a/funasr/utils/load_utils.py b/funasr/utils/load_utils.py
index 4849408..8ff7115 100644
--- a/funasr/utils/load_utils.py
+++ b/funasr/utils/load_utils.py
@@ -19,18 +19,19 @@
def is_ffmpeg_installed():
try:
- # 灏濊瘯杩愯ffmpeg鍛戒护骞惰幏鍙栧叾鐗堟湰淇℃伅
output = subprocess.check_output(['ffmpeg', '-version'], stderr=subprocess.STDOUT)
return 'ffmpeg version' in output.decode('utf-8')
except (subprocess.CalledProcessError, FileNotFoundError):
- # 鑻ヨ繍琛宖fmpeg鍛戒护澶辫触锛屽垯璁や负ffmpeg鏈畨瑁�
return False
use_ffmpeg=False
if is_ffmpeg_installed():
use_ffmpeg = True
else:
- print("Notice: ffmpeg is not installed. torchaudio is used to load audio")
+ print("Notice: ffmpeg is not installed. torchaudio is used to load audio\n"
+ "If you want to use ffmpeg backend to load audio, please install it by:"
+ "\n\tsudo apt install ffmpeg # ubuntu"
+ "\n\t# brew install ffmpeg # mac")
def load_audio_text_image_video(data_or_path_or_list, fs: int = 16000, audio_fs: int = 16000, data_type="sound", tokenizer=None, **kwargs):
if isinstance(data_or_path_or_list, (list, tuple)):
@@ -50,13 +51,20 @@
if isinstance(data_or_path_or_list, str) and os.path.exists(data_or_path_or_list): # local file
if data_type is None or data_type == "sound":
- if use_ffmpeg:
- data_or_path_or_list = _load_audio_ffmpeg(data_or_path_or_list, sr=fs)
- data_or_path_or_list = torch.from_numpy(data_or_path_or_list).squeeze() # [n_samples,]
- else:
+ # if use_ffmpeg:
+ # data_or_path_or_list = _load_audio_ffmpeg(data_or_path_or_list, sr=fs)
+ # data_or_path_or_list = torch.from_numpy(data_or_path_or_list).squeeze() # [n_samples,]
+ # else:
+ # data_or_path_or_list, audio_fs = torchaudio.load(data_or_path_or_list)
+ # if kwargs.get("reduce_channels", True):
+ # data_or_path_or_list = data_or_path_or_list.mean(0)
+ try:
data_or_path_or_list, audio_fs = torchaudio.load(data_or_path_or_list)
if kwargs.get("reduce_channels", True):
data_or_path_or_list = data_or_path_or_list.mean(0)
+ except:
+ data_or_path_or_list = _load_audio_ffmpeg(data_or_path_or_list, sr=fs)
+ data_or_path_or_list = torch.from_numpy(data_or_path_or_list).squeeze() # [n_samples,]
elif data_type == "text" and tokenizer is not None:
data_or_path_or_list = tokenizer.encode(data_or_path_or_list)
elif data_type == "image": # undo
--
Gitblit v1.9.1