Update app.py

app.py

@@ -3,7 +3,8 @@ import gradio as gr
 import requests
 import pandas as pd
 import base64
-import ollama
+
+from dotenv import load_dotenv
 from typing import TypedDict, Annotated
 from langchain_core.messages import AnyMessage, SystemMessage
 from langchain_community.tools import DuckDuckGoSearchRun
@@ -13,10 +14,18 @@ from langgraph.graph import START, StateGraph
 from langgraph.graph.message import add_messages
 from langgraph.prebuilt import ToolNode, tools_condition
 
+from huggingface_hub import login, InferenceClient
+
 # (Keep Constants as is)
 # --- Constants ---
 DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
 
+load_dotenv()
+login(os.getenv("HUGGINGFACEHUB_API_TOKEN"))
+
+model_id = "google/gemma-3-4b-it"
+
+client = InferenceClient(model=model_id)
 
 def get_file_path(task_id: str, question) -> str:
     """Retrieves reference file path."""
@@ -81,13 +90,12 @@ extract_text_tool = Tool(
     description="Retrieves text from an image."
 )
 
-# Generate the chat interface, including the tools
-# print("Loading LLM...")
-# ollama.pull('gemma3:4b')
-from transformers import AutoProcessor, AutoModelForImageTextToText
-
-processor = AutoProcessor.from_pretrained("google/gemma-3-4b-it")
-model = AutoModelForImageTextToText.from_pretrained("google/gemma-3-4b-it")
+model = HfApiModel(
+max_tokens=2096,
+temperature=0.5,
+model_id='Qwen/Qwen2.5-Coder-32B-Instruct',# it is possible that this model may be overloaded
+custom_role_conversions=None,
+)
 
 chat = ChatOllama(model=model, verbose=True)
 print(f"Model {chat.model} downloaded!")