Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
@@ -4,6 +4,7 @@ import os
|
|
4 |
import zipfile
|
5 |
import requests
|
6 |
from TTS.api import TTS
|
|
|
7 |
|
8 |
# Set environment variable
|
9 |
os.environ["COQUI_TOS_AGREED"] = "1"
|
@@ -11,6 +12,7 @@ os.environ["COQUI_TOS_AGREED"] = "1"
|
|
11 |
# Define constants
|
12 |
MODEL_PATH = "tts_models/multilingual/multi-dataset/xtts_v2"
|
13 |
LANGUAGES = ["en", "es", "fr", "de", "it", "pt", "pl", "tr", "ru", "nl", "cs", "ar", "zh-cn", "ja", "hu", "ko", "hi"]
|
|
|
14 |
|
15 |
# Automatically detect and use GPU if available
|
16 |
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
|
@@ -19,14 +21,35 @@ print(f"Using device: {device}")
|
|
19 |
# Load TTS model
|
20 |
tts = TTS(MODEL_PATH).to(device)
|
21 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
22 |
def clone(text, url, language):
|
23 |
-
# Download
|
24 |
response = requests.get(url)
|
25 |
with open("temp.zip", "wb") as f:
|
26 |
f.write(response.content)
|
|
|
|
|
27 |
with zipfile.ZipFile("temp.zip", "r") as zip_ref:
|
28 |
-
zip_ref.
|
29 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
30 |
|
31 |
# Generate audio using TTS model
|
32 |
tts.tts_to_file(text=text, speaker_wav=audio_file, language=language, file_path="./output.wav")
|
|
|
4 |
import zipfile
|
5 |
import requests
|
6 |
from TTS.api import TTS
|
7 |
+
from pydub import AudioSegment
|
8 |
|
9 |
# Set environment variable
|
10 |
os.environ["COQUI_TOS_AGREED"] = "1"
|
|
|
12 |
# Define constants
|
13 |
MODEL_PATH = "tts_models/multilingual/multi-dataset/xtts_v2"
|
14 |
LANGUAGES = ["en", "es", "fr", "de", "it", "pt", "pl", "tr", "ru", "nl", "cs", "ar", "zh-cn", "ja", "hu", "ko", "hi"]
|
15 |
+
AUDIO_FORMATS = [".wav", ".mp3", ".flac", ".mp4"]
|
16 |
|
17 |
# Automatically detect and use GPU if available
|
18 |
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
|
|
|
21 |
# Load TTS model
|
22 |
tts = TTS(MODEL_PATH).to(device)
|
23 |
|
24 |
+
def convert_to_wav(audio_file):
|
25 |
+
# Extract file extension
|
26 |
+
file_extension = os.path.splitext(audio_file)[-1].lower()
|
27 |
+
|
28 |
+
# Convert audio file to .wav format
|
29 |
+
if file_extension != ".wav":
|
30 |
+
audio = AudioSegment.from_file(audio_file)
|
31 |
+
audio.export("temp.wav", format="wav")
|
32 |
+
os.remove(audio_file)
|
33 |
+
audio_file = "temp.wav"
|
34 |
+
|
35 |
+
return audio_file
|
36 |
+
|
37 |
def clone(text, url, language):
|
38 |
+
# Download zip file
|
39 |
response = requests.get(url)
|
40 |
with open("temp.zip", "wb") as f:
|
41 |
f.write(response.content)
|
42 |
+
|
43 |
+
# Extract audio file from zip archive
|
44 |
with zipfile.ZipFile("temp.zip", "r") as zip_ref:
|
45 |
+
for file in zip_ref.namelist():
|
46 |
+
if os.path.splitext(file)[-1].lower() in AUDIO_FORMATS:
|
47 |
+
zip_ref.extract(file, ".")
|
48 |
+
audio_file = file
|
49 |
+
break
|
50 |
+
|
51 |
+
# Convert audio file to .wav format
|
52 |
+
audio_file = convert_to_wav(audio_file)
|
53 |
|
54 |
# Generate audio using TTS model
|
55 |
tts.tts_to_file(text=text, speaker_wav=audio_file, language=language, file_path="./output.wav")
|