Update main.py

main.py

@@ -1,55 +1,23 @@
-import gradio as gr
-from pymongo import MongoClient
-import pandas as pd
+from fastapi import FastAPI
+from pydantic import BaseModel
 import os
+app = FastAPI()
 
-
-secret_key = os.getenv('huggingface_key')
-print(f"Secret Key: {secret_key}")
-
-
+model = load_model()
 
 client = MongoClient('mongodb://localhost:27017/')
-db = client['EgitimDatabase']
-collection = db['test']
+db = client['']
+collection = db['text']
 
-def train_model(filtered_data):
-   
-    
-    model_response = {
-        'status': 'success',
-        'message': 'Model trained successfully!',
-        'accuracy': 0.95,  # Örnek doğruluk değeri
-        'data_size': len(filtered_data)
-    }
-    return model_response
 
-# Gradio uygulaması için fonksiyon
-def train_model_gradio(title,keywords,subheadings):
-   
-    query = {
-        'title': {'$in': title},
-        'category': {'$in': keywords.split(',')},
-        'subheadings': {'$in': subheadings.split(',')}
-    }
-    filtered_data = list(collection.find(query))
 
-    
-    response = train_model(filtered_data)
-    return response
-
-# Gradio arayüzü
-iface = gr.Interface(
-    fn=train_model_gradio,
-    inputs=[
-        gr.Textbox(label="Title"),
-        gr.Textbox(label="Keywords (comma-separated)"),
-        gr.Textbox(label="Subheadings (comma-separated)")
-    ],
-    outputs="json",
-    title="Model Training Interface",
-    description="Enter the titles, categories, subcategories, and subheadings to filter the data and train the model."
-)
+class InputData(BaseModel):
+    title: str
+    keywords: str
+    subheadings: str
 
-if __name__ == "__main__":
-    iface.launch()
+@app.post("/generate/")
+async def generate(input_data: InputData):
+    
+     result = model.predict(input_data)
+    return {"result": "generated_output"}