Create app.py

app.py

@@ -0,0 +1,29 @@
+import gradio as gr
+import torch
+from transformers import Tacotron2Processor, Tacotron2ForTextToSpeech, WaveglowForVocoder
+import soundfile as sf
+
+device = "cuda" if torch.cuda.is_available() else "cpu"
+
+processor = Tacotron2Processor.from_pretrained("facebook/tts-transformer-multilingual")
+model = Tacotron2ForTextToSpeech.from_pretrained("facebook/tts-transformer-multilingual").to(device)
+vocoder = WaveglowForVocoder.from_pretrained("facebook/tts-transformer-multilingual").to(device)
+
+def text_to_speech(text):
+    inputs = processor(text=text, return_tensors="pt")
+    mel_outputs = model(**inputs.to(device)).mel_outputs
+    with torch.no_grad():
+        waveform = vocoder(mel_outputs)
+    waveform = waveform.cpu().numpy()
+    sf.write("output.wav", waveform.T, samplerate=22050)
+    return "output.wav"
+
+iface = gr.Interface(
+    fn=text_to_speech,
+    inputs=gr.Textbox(lines=2, placeholder="متن خود را اینجا وارد کنید..."),
+    outputs="audio",
+    title="تبدیل متن فارسی به صدا",
+    description="متن فارسی خود را وارد کنید تا صدای آن را دریافت کنید.",
+)
+
+iface.launch()