Update app.py

app.py

@@ -1,23 +1,31 @@
 import gradio as gr
-from bark import SAMPLE_RATE, generate_audio, preload_models
-from IPython.display import Audio
+from fairseq.checkpoint_utils import load_model_ensemble_and_task_from_hf_hub
+from fairseq.models.text_to_speech.hub_interface import TTSHubInterface
+import numpy as np
 
-# Preload models if needed
-preload_models()
+def load_tts_model():
+    models, cfg, task = load_model_ensemble_and_task_from_hf_hub(
+        "facebook/tts_transformer-zh-cv7_css10",
+        arg_overrides={"vocoder": "hifigan", "fp16": False}
+    )
+    model = models[0]
+    TTSHubInterface.update_cfg_with_data_cfg(cfg, task.data_cfg)
+    generator = task.build_generator(model, cfg)
+    return task, model, generator
 
-def create_audio(text):
-    audio_array = generate_audio(text)
-    return audio_array, SAMPLE_RATE
+task, model, generator = load_tts_model()
+
+def synthesize_text(text):
+    sample = TTSHubInterface.get_model_input(task, text)
+    wav, rate = TTSHubInterface.get_prediction(task, model, generator, sample)
+    return np.array(wav), rate
 
 iface = gr.Interface(
-    fn=create_audio,
-    inputs=gr.inputs.Textbox(lines=5, placeholder="Digite seu texto aqui..."),
-    outputs=[
-        gr.outputs.Audio(type="numpy", label="Audio Gerado"),
-        gr.outputs.Textbox(label="Taxa de Amostragem")
-    ],
-    title="Gerador de Áudio com IA",
-    description="Digite um texto para gerar áudio usando IA."
+    fn=synthesize_text,
+    inputs=gr.inputs.Textbox(lines=2, placeholder="Enter Chinese text here..."),
+    outputs=gr.outputs.Audio(label="Synthesized Speech"),
+    title="Text to Speech Synthesis",
+    description="A simple text-to-speech app using Fairseq TTS model."
 )
 
 iface.launch()