Update app.py

app.py

@@ -1,15 +1,10 @@
 import os
-import shutil
-import tempfile
 import torch
 import spaces
 import gradio as gr
-from fastapi import UploadFile, File, Form, HTTPException
-from fastapi.middleware.cors import CORSMiddleware
-from fastapi.responses import JSONResponse, PlainTextResponse
 from transformers import pipeline
 
-# 1. 初始化模型
+# 1. 初始化 Whisper 模型
 MODEL_NAME = "openai/whisper-small"
 device = "cuda" if torch.cuda.is_available() else "cpu"
 
@@ -20,11 +15,14 @@ pipe = pipeline(
     device=device
 )
 
-# 2. 顶层 ZeroGPU 函数
+# 2. ZeroGPU 动态调用函数
 @spaces.GPU
 def run_whisper(audio_path: str, target_language: str = None, is_translate: bool = False):
+    if not audio_path:
+        return "请上传或录制音频文件！"
+        
     generate_kwargs = {}
-    if target_language:
+    if target_language and target_language != "auto":
         generate_kwargs["language"] = target_language
 
     if is_translate:
@@ -34,64 +32,46 @@ def run_whisper(audio_path: str, target_language: str = None, is_translate: bool
     result = pipe(audio_path, generate_kwargs=generate_kwargs)
     return result["text"]
 
-# 3. Gradio 界面定义
-def gradio_predict(audio_path):
-    if audio_path is None:
-        return "请上传音频文件！"
-    return run_whisper(audio_path)
+# 3. Gradio 交互逻辑
+def gradio_predict(audio_path, language, is_translate):
+    return run_whisper(audio_path, target_language=language, is_translate=is_translate)
 
-demo = gr.Interface(
-    fn=gradio_predict,
-    inputs=gr.Audio(sources=["microphone", "upload"], type="filepath", label="输入音频"),
-    outputs=gr.Textbox(label="识别结果"),
-    title="Whisper API Node"
-)
-
-# 4. 获取 FastAPI 实例并添加 CORS 跨域支持
-app = demo.app
-
-app.add_middleware(
-    CORSMiddleware,
-    allow_origins=["*"],
-    allow_credentials=True,
-    allow_methods=["*"],
-    allow_headers=["*"],
-)
-
-async def process_audio(file: UploadFile, response_format: str, language: str, is_translate: bool):
-    suffix = os.path.splitext(file.filename)[1] or ".wav"
-    with tempfile.NamedTemporaryFile(delete=False, suffix=suffix) as temp_file:
-        shutil.copyfileobj(file.file, temp_file)
-        temp_path = temp_file.name
-
-    try:
-        text = run_whisper(temp_path, target_language=language, is_translate=is_translate)
-    except Exception as e:
-        raise HTTPException(status_code=500, detail=str(e))
-    finally:
-        if os.path.exists(temp_path):
-            os.remove(temp_path)
-
-    if response_format in ["text", "vtt", "srt"]:
-        return PlainTextResponse(text)
-    return JSONResponse(content={"text": text})
-
-@app.post("/v1/audio/transcriptions")
-async def transcribe_api(
-    file: UploadFile = File(...),
-    model: str = Form("whisper-1"),
-    language: str = Form(None),
-    response_format: str = Form("json")
-):
-    return await process_audio(file, response_format, language, is_translate=False)
+# 4. 构建 Gradio 界面
+with gr.Blocks(title="Whisper 语音识别与翻译") as demo:
+    gr.Markdown("## 🎙️ Whisper 语音识别与翻译工具")
+    
+    with gr.Row():
+        with gr.Column():
+            audio_input = gr.Audio(
+                sources=["microphone", "upload"], 
+                type="filepath", 
+                label="上传或录制音频"
+            )
+            
+            # 增加常用语言选择与翻译开关选项
+            language_dropdown = gr.Dropdown(
+                choices=["auto", "chinese", "english", "japanese", "korean", "cantonese"],
+                value="auto",
+                label="指定源语言 (默认自动识别)"
+            )
+            
+            translate_checkbox = gr.Checkbox(
+                label="翻译为英文 (Task: Translate to English)", 
+                value=False
+            )
+            
+            submit_btn = gr.Button("开始识别 / 翻译", variant="primary")
+            
+        with gr.Column():
+            text_output = gr.Textbox(label="识别 / 翻译结果", lines=10)
 
-@app.post("/v1/audio/translations")
-async def translate_api(
-    file: UploadFile = File(...),
-    model: str = Form("whisper-1"),
-    response_format: str = Form("json")
-):
-    return await process_audio(file, response_format, language="english", is_translate=True)
+    # 绑定事件
+    submit_btn.click(
+        fn=gradio_predict,
+        inputs=[audio_input, language_dropdown, translate_checkbox],
+        outputs=text_output
+    )
 
 # 5. 启动服务
-demo.launch()
+if __name__ == "__main__":
+    demo.launch()