Upload app0.py with huggingface_hub

app0.py

@@ -0,0 +1,110 @@
+import os
+import shutil
+import threading
+import time
+import sys
+import torch
+
+from huggingface_hub import snapshot_download
+
+current_dir = os.path.dirname(os.path.abspath(__file__))
+sys.path.append(current_dir)
+sys.path.append(os.path.join(current_dir, "indextts"))
+
+import gradio as gr
+from indextts.infer import IndexTTS
+from tools.i18n.i18n import I18nAuto
+
+# 设置多语言
+i18n = I18nAuto(language="en")
+
+# 下载模型
+MODE = 'local'
+snapshot_download("IndexTeam/IndexTTS-1.5", local_dir="checkpoints")
+
+# 自动选择设备：优先 GPU，没有就 CPU
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+print(f"🔥 Using device: {device}")
+
+# 初始化 TTS (不传 device)
+tts = IndexTTS(model_dir="checkpoints", cfg_path="checkpoints/config.yaml")
+
+# 如果 IndexTTS 支持 to() 或 to_device()，切换设备
+if hasattr(tts, "to"):
+    tts.to(device)
+elif hasattr(tts, "to_device"):
+    tts.to_device(device)
+else:
+    print("⚠️ IndexTTS 没有 to()/to_device() 方法，可能内部已自动处理设备。")
+
+# 确保必要的目录存在
+os.makedirs("outputs/tasks", exist_ok=True)
+os.makedirs("prompts", exist_ok=True)
+
+# 推理函数
+def infer(voice, text, output_path=None):
+    if not tts:
+        raise Exception("Model not loaded")
+    if not output_path:
+        output_path = os.path.join("outputs", f"spk_{int(time.time())}.wav")
+    tts.infer(voice, text, output_path)
+    return output_path
+
+# API 调用
+def tts_api(voice, text):
+    try:
+        output_path = infer(voice, text)
+        with open(output_path, "rb") as f:
+            audio_bytes = f.read()
+        return (200, {}, audio_bytes)
+    except Exception as e:
+        return (500, {"error": str(e)}, None)
+
+# 单次生成
+def gen_single(prompt, text):
+    output_path = infer(prompt, text)
+    return gr.update(value=output_path, visible=True)
+
+# 上传后启用按钮
+def update_prompt_audio():
+    update_button = gr.update(interactive=True)
+    return update_button
+
+# Gradio 界面
+with gr.Blocks() as demo:
+    mutex = threading.Lock()
+    gr.HTML(f'''
+    <h2><center>IndexTTS WebUI</center></h2>
+    <p align="center">当前设备: <b>{device}</b></p>
+    ''')
+    with gr.Tab("音频生成"):
+        with gr.Row():
+            os.makedirs("prompts", exist_ok=True)
+            prompt_audio = gr.Audio(
+                label="请上传参考音频",
+                key="prompt_audio",
+                sources=["upload", "microphone"],
+                type="filepath"
+            )
+            prompt_list = os.listdir("prompts")
+            default = ''
+            if prompt_list:
+                default = prompt_list[0]
+            input_text_single = gr.Textbox(label="请输入目标文本", key="input_text_single")
+            gen_button = gr.Button("生成语音", key="gen_button", interactive=True)
+            output_audio = gr.Audio(label="生成结果", visible=False, key="output_audio")
+
+    prompt_audio.upload(update_prompt_audio,
+                         inputs=[],
+                         outputs=[gen_button])
+
+    gen_button.click(gen_single,
+                     inputs=[prompt_audio, input_text_single],
+                     outputs=[output_audio])
+
+def main():
+    tts.load_normalizer()
+    demo.launch(server_name="0.0.0.0", server_port=7860)
+
+if __name__ == "__main__":
+    main()