example_usage.py

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
TIGeR Model Usage Examples
Demonstrates how to use the fine-tuned spatial reasoning model
"""

import os
import sys
import json
import re
from typing import List, Dict, Any, Optional
from llamafactory.chat.chat_model import ChatModel

def load_model(config_file: str):
    """Load model using LLaMA-Factory ChatModel"""
    print(f"Loading model with config: {config_file}")
    
    try:
        # Simulate command line arguments
        original_argv = sys.argv.copy()
        sys.argv = [sys.argv[0], config_file]
        
        try:
            chat_model = ChatModel()
            print("✅ Model loaded successfully!")
            return chat_model
        finally:
            # Restore original command line arguments
            sys.argv = original_argv
            
    except Exception as e:
        print(f"❌ Model loading failed: {e}")
        return None

def single_inference_demo(chat_model):
    """Single image inference demonstration"""
    print("\n" + "="*50)
    print("Single Image Inference Demo")
    print("="*50)
    
    # Image path - replace with your actual image path
    image_paths = [
        "/path/to/your/image.jpg"  # Replace with actual image path
    ]
    
    # Question - using the same format as TIGeR
    question = "Two points are circled on the image, labeled by A and B beside each circle. Which point is closer to the camera? Select from the following choices.\n(A) A is closer\n(B) B is closer"
    
    try:
        print(f"📷 Loading image: {image_paths[0]}")
        print(f"🔍 Question: {question}")
        
        # Prepare messages in the format expected by ChatModel
        messages = [
            {
                "role": "user",
                "content": question
            }
        ]
        
        # Get model response
        response = chat_model.chat(messages, images=image_paths)
        assistant_texts = []
        
        for resp in response:
            try:
                assistant_texts.append(resp.response_text)
            except Exception:
                assistant_texts.append(str(resp))
        
        response_text = "\n".join(assistant_texts)
        print(f"💡 Answer: {response_text}")
                
    except FileNotFoundError:
        print("❌ Image file not found, please provide correct image path")
    except Exception as e:
        print(f"❌ Error occurred during processing: {e}")

def main():
    """Main function"""
    print("🚀 TIGeR Model Usage Examples")
    print("="*60)
    
    # Configuration file path - using the config file in the same directory
    config_file = "glm4v_tisr_full_inference.yaml"
    
    # Load model
    chat_model = load_model(config_file)
    if chat_model is None:
        print("❌ Failed to load model. Please check the config file path.")
        return
    
    # Run single inference demo
    single_inference_demo(chat_model)
    
    print("\n" + "="*60)
    print("✅ Demo completed!")
    print("="*60)

if __name__ == "__main__":
    main()