Update app.py
Browse files
app.py
CHANGED
|
@@ -1,49 +1,24 @@
|
|
| 1 |
-
import os
|
| 2 |
-
os.environ["COQUI_TOS_AGREED"] = "1"
|
| 3 |
-
|
| 4 |
-
import torch
|
| 5 |
-
from torch.serialization import add_safe_globals
|
| 6 |
-
|
| 7 |
-
from TTS.tts.configs.xtts_config import XttsConfig
|
| 8 |
-
from TTS.tts.models.xtts import XttsAudioConfig
|
| 9 |
-
from TTS.config.shared_configs import BaseDatasetConfig # NEW ADD
|
| 10 |
-
|
| 11 |
-
add_safe_globals([XttsConfig, XttsAudioConfig, BaseDatasetConfig])
|
| 12 |
-
|
| 13 |
from TTS.api import TTS
|
| 14 |
import gradio as gr
|
|
|
|
|
|
|
|
|
|
|
|
|
| 15 |
|
| 16 |
-
tts = TTS(model_name="tts_models/
|
| 17 |
tts.to("cuda" if torch.cuda.is_available() else "cpu")
|
| 18 |
|
| 19 |
-
def synthesize(text
|
| 20 |
output_path = "output.wav"
|
| 21 |
-
|
| 22 |
-
|
| 23 |
-
|
| 24 |
-
speaker_path = "temp_speaker.wav"
|
| 25 |
-
speaker_wav.save(speaker_path)
|
| 26 |
-
|
| 27 |
-
tts.tts_to_file(
|
| 28 |
-
text=text,
|
| 29 |
-
speaker_wav=speaker_path,
|
| 30 |
-
language="en",
|
| 31 |
-
file_path=output_path
|
| 32 |
-
)
|
| 33 |
-
return "Done. Listen below.", output_path
|
| 34 |
|
| 35 |
iface = gr.Interface(
|
| 36 |
fn=synthesize,
|
| 37 |
-
inputs=
|
| 38 |
-
|
| 39 |
-
|
| 40 |
-
|
| 41 |
-
outputs=[
|
| 42 |
-
gr.Textbox(label="Status"),
|
| 43 |
-
gr.Audio(label="Result Audio")
|
| 44 |
-
],
|
| 45 |
-
title="XTTS‑V2 Horror Whisperer 👻",
|
| 46 |
-
description="Upload a voice and make it say creepy stuff. XTTS‑v2 powered, GPU-juiced. 🧟♂️💀"
|
| 47 |
)
|
| 48 |
|
| 49 |
if __name__ == "__main__":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
from TTS.api import TTS
|
| 2 |
import gradio as gr
|
| 3 |
+
import torch
|
| 4 |
+
import os
|
| 5 |
+
|
| 6 |
+
os.environ["CUDA_VISIBLE_DEVICES"] = "0" # Use GPU if available
|
| 7 |
|
| 8 |
+
tts = TTS(model_name="tts_models/en/ljspeech/tacotron2-DDC")
|
| 9 |
tts.to("cuda" if torch.cuda.is_available() else "cpu")
|
| 10 |
|
| 11 |
+
def synthesize(text):
|
| 12 |
output_path = "output.wav"
|
| 13 |
+
tts.tts_to_file(text=text, file_path=output_path)
|
| 14 |
+
return output_path
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 15 |
|
| 16 |
iface = gr.Interface(
|
| 17 |
fn=synthesize,
|
| 18 |
+
inputs=gr.Textbox(lines=3, label="Enter text to synthesize"),
|
| 19 |
+
outputs=gr.Audio(type="filepath", label="Generated Speech"),
|
| 20 |
+
title="Mozilla Tacotron2 TTS - Smooth & Clean",
|
| 21 |
+
description="Simple text-to-speech using Mozilla's Tacotron2-DDC model. No voice samples required."
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 22 |
)
|
| 23 |
|
| 24 |
if __name__ == "__main__":
|