Update app.py

app.py

@@ -1,90 +1,31 @@
-import os
-import torch
-import scipy.io.wavfile
+from flask import Flask, request, jsonify
+from api import UrduWhisper
 import uuid
-from fastapi import FastAPI, HTTPException, Header
-from fastapi.responses import FileResponse, JSONResponse
-from fastapi.middleware.cors import CORSMiddleware
-from diffusers import AudioLDM2Pipeline
-
-app = FastAPI()
-
-# --- CORS Permissions ---
-app.add_middleware(
-    CORSMiddleware,
-    allow_origins=["*"],
-    allow_credentials=True,
-    allow_methods=["*"],
-    allow_headers=["*"],
-)
-
-OUTPUT_DIR = "/tmp"
-os.makedirs(OUTPUT_DIR, exist_ok=True)
-API_KEY = os.getenv("API_KEY", "MySecretPassword123")
-
-# --- MODEL LOADING (Startup par 1-2 min lega) ---
-print("⏳ Loading Music AI Model (AudioLDM-2 Lite)...")
-try:
-    pipe = AudioLDM2Pipeline.from_pretrained(
-        "cvssp/audioldm2-lite",  # Lite version (CPU friendly)
-        torch_dtype=torch.float32 # CPU par float32 best hai
-    )
-    # Speed Optimization: CPU Offload na karein, direct CPU use karein
-    pipe.to("cpu") 
-    print("✅ Music Model Loaded Successfully!")
-except Exception as e:
-    print(f"❌ Error Loading Model: {e}")
-    pipe = None
-
-@app.get("/")
-def home():
-    if pipe is None:
-        return {"status": "Error", "message": "Model failed to load. Check Logs."}
-    return {"status": "Online", "message": "AI Music Generator Ready"}
-
-@app.post("/generate")
-async def generate_music(
-    prompt: str,
-    duration: int = 5,  # Seconds (Free tier ke liye 5-10s best hai)
-    x_api_key: str = Header(None)
-):
-    # 1. Security Check
-    if x_api_key != API_KEY:
-        raise HTTPException(status_code=401, detail="Invalid API Key")
+import os
 
-    if pipe is None:
-        raise HTTPException(status_code=500, detail="Model is not loaded")
+app = Flask(__name__)
+model = UrduWhisper()
 
-    # Limit Duration to avoid crashing free tier
-    if duration > 10:
-        duration = 10 
+UPLOAD = "uploads"
+os.makedirs(UPLOAD, exist_ok=True)
 
-    filename = f"{uuid.uuid4()}.wav"
-    filepath = os.path.join(OUTPUT_DIR, filename)
+@app.route("/transcribe", methods=["POST"])
+def transcribe_audio():
+    if "file" not in request.files:
+        return jsonify({"error": "No file uploaded"}), 400
+    
+    file = request.files["file"]
+    filename = f"{UPLOAD}/{uuid.uuid4()}.wav"
+    file.save(filename)
 
-    try:
-        print(f"🎵 Generating Music: '{prompt}' ({duration}s)...")
-        
-        # 2. GENERATION LOGIC
-        # num_inference_steps=10 rakha hai taakay jaldi banay (Quality thori kam hogi par speed achi hogi)
-        # Agar quality achi chahiye to 20 kar dein (Lekin time double lagega)
-        audio = pipe(
-            prompt,
-            num_inference_steps=10, 
-            audio_length_in_s=duration,
-            negative_prompt="low quality, average quality, noise"
-        ).audios[0]
+    text = model.transcribe(filename)
+    os.remove(filename)
 
-        # 3. Save File
-        scipy.io.wavfile.write(filepath, rate=16000, data=audio)
-        
-        print("✅ Music Saved!")
-        return FileResponse(filepath, media_type="audio/wav", filename="music.wav")
+    return jsonify({"text": text})
 
-    except Exception as e:
-        print(f"❌ Error: {e}")
-        return JSONResponse(status_code=500, content={"error": str(e)})
+@app.route("/")
+def home():
+    return {"message": "Custom Urdu Whisper API Running!"}
 
 if __name__ == "__main__":
-    import uvicorn
-    uvicorn.run(app, host="0.0.0.0", port=7860)
+    app.run(host="0.0.0.0", port=5000)