Update app.py

app.py

@@ -1,29 +1,48 @@
 import gradio as gr
 from transformers import pipeline
-import fastapi
-from gradio import fastapi as gr_api
+from fastapi import FastAPI
+from gradio import routes
+import uvicorn
 
+# Load the model (this runs only once!)
 generator = pipeline("text2text-generation", model="LahiruProjects/recipe-generator-flan-t5")
 
+# Function to generate recipe steps
 def generate_recipe(name, ingredients, calories, time):
     prompt = f"""Create a step-by-step recipe for "{name}" using these ingredients: {', '.join(ingredients.split(','))}.
     Keep it under {calories} calories and make sure it's ready in less than {time} minutes."""
     result = generator(prompt)
     return result[0]["generated_text"]
 
-# Using gr.Blocks() for a custom interface
-with gr.Blocks() as demo:
-    with gr.Row():
-        name = gr.Textbox(label="Recipe Name")
-        ingredients = gr.Textbox(label="Ingredients (comma-separated)")
-        calories = gr.Number(label="Max Calories", value=400)
-        time = gr.Number(label="Max Cooking Time (minutes)", value=30)
-    output = gr.Textbox()
-    demo.add(name, ingredients, calories, time, output)
+# Gradio interface
+iface = gr.Interface(
+    fn=generate_recipe,
+    inputs=[
+        gr.Textbox(label="Recipe Name"),
+        gr.Textbox(label="Ingredients (comma-separated)"),
+        gr.Number(label="Max Calories", value=400),
+        gr.Number(label="Max Cooking Time (minutes)", value=30)
+    ],
+    outputs="text",
+    title="🍳 Recipe Generator (FLAN-T5)",
+    description="Generate a step-by-step recipe based on ingredients, calorie limit, and time"
+)
 
-# Set up FastAPI app to expose the REST API
-app = fastapi.FastAPI()
-app = gr_api.FastAPI(app, demo)
+# FastAPI integration
+app = FastAPI()
 
-gr_api.add_api_route(app, "/api/predict", demo, methods=["POST"])
+# Define the API route using Gradio interface
+@app.post("/api/predict")
+async def predict(data: list):
+    inputs = data[0]
+    name = inputs[0]
+    ingredients = inputs[1]
+    calories = inputs[2]
+    time = inputs[3]
+    result = generate_recipe(name, ingredients, calories, time)
+    return {"data": [result]}
 
+# Serve the Gradio interface and FastAPI app using Uvicorn (locally for testing)
+if __name__ == "__main__":
+    iface.launch(server_name="0.0.0.0", server_port=7860)  # This will launch the Gradio app
+    uvicorn.run(app, host="0.0.0.0", port=8000)  # FastAPI app will run on port 8000