scripts/enhanced_test_l4.py

import requests
import json
import time

# Test Turkish Medical Model API - L4 Optimized
BASE_URL = "https://conquerorr000-turkish-medical-model-api.hf.space"

def test_endpoint(endpoint, method="GET", data=None):
    """Test an API endpoint"""
    url = f"{BASE_URL}{endpoint}"
    
    try:
        start_time = time.time()
        
        if method == "POST":
            response = requests.post(url, json=data, timeout=60)
        else:
            response = requests.get(url, timeout=30)
        
        response_time = time.time() - start_time
        
        print(f"\n{'='*50}")
        print(f" {method} {endpoint}")
        print(f" Status Code: {response.status_code}")
        print(f"⏱️ Response Time: {response_time:.2f}s")
        
        if response.status_code == 200:
            result = response.json()
            print(f" Response:")
            print(json.dumps(result, indent=2, ensure_ascii=False))
            return result
        else:
            print(f"❌ Error: {response.status_code}")
            print(f"📋 Response: {response.text}")
            return None
        
    except Exception as e:
        print(f" Error testing {endpoint}: {e}")
        return None

def test_medical_conversation():
    """Test a full medical conversation"""
    print(f"\n{'='*50}")
    print(" MEDICAL CONVERSATION TEST")
    print(f"{'='*50}")
    
    # Conversation messages
    conversation_data = {
        "messages": [
            {"role": "user", "content": "Merhaba doktor, 3 gündür ateşim var."},
            {"role": "assistant", "content": "Merhaba. Ateşinizin kaç derece olduğunu ölçtünüz mü?"},
            {"role": "user", "content": "Evet, 38.5 derece civarında. Ayrıca boğazım da ağrıyor."}
        ],
        "max_tokens": 150,
        "temperature": 0.7
    }
    
    return test_endpoint("/conversation", "POST", conversation_data)

def performance_benchmark():
    """Run performance benchmark"""
    print(f"\n{'='*50}")
    print(" PERFORMANCE BENCHMARK")
    print(f"{'='*50}")
    
    test_cases = [
        "Başım ağrıyor ve mide bulantım var.",
        "2 haftadır öksürüyorum ve balgamım var.",
        "Dizlerimde ağrı var, özellikle sabahları.",
        "Uykusuzluk problemi yaşıyorum, ne yapmalıyım?",
        "Alerjik reaksiyonum olabilir, derimde kaşıntı var."
    ]
    
    total_time = 0
    successful_tests = 0
    
    for i, test_case in enumerate(test_cases, 1):
        print(f"\n Test Case {i}: {test_case}")
        
        chat_data = {
            "message": test_case,
            "max_tokens": 120,
            "temperature": 0.7
        }
        
        result = test_endpoint("/chat", "POST", chat_data)
        
        if result and result.get("response"):
            generation_time = result.get("generation_time", 0)
            total_time += generation_time
            successful_tests += 1
            print(f" Generation Time: {generation_time:.2f}s")
            print(f" Response Quality: {'Clean' if len(result['response']) > 20 and 'Hasta' not in result['response'] else 'Needs Improvement'}")
        else:
            print(" Test failed")
    
    if successful_tests > 0:
        avg_time = total_time / successful_tests
        print(f"\n BENCHMARK RESULTS:")
        print(f" Successful Tests: {successful_tests}/{len(test_cases)}")
        print(f" Average Generation Time: {avg_time:.2f}s")
        print(f" Performance Rating: {'Excellent' if avg_time < 5 else 'Good' if avg_time < 7 else 'Needs Optimization'}")

# Main test execution
print(" Turkish Medical Model API - L4 Performance Test")
print(" GPU: Nvidia L4 24GB VRAM")
print(" Mode: FP16 Full Precision + Flash Attention")

# 1. Health Check
health_result = test_endpoint("/health")

if health_result and health_result.get("model_loaded"):
    print("\n Model is loaded and ready!")
    
    # 2. Memory Status
    test_endpoint("/memory-status")
    
    # 3. Debug Info
    test_endpoint("/debug")
    
    # 4. Quick Test
    test_endpoint("/test")
    
    # 5. Medical Conversation Test
    test_medical_conversation()
    
    # 6. Performance Benchmark
    performance_benchmark()
    
else:
    print("\n Model not ready, skipping performance tests")
    print("ℹ Wait for model to load and try again")

print(f"\n{'='*50}")
print(" L4 Performance Test Completed!")