testign promtps

test_large_context.py

@@ -0,0 +1,298 @@
+#!/usr/bin/env python
+"""
+Test AI with LARGE INPUT (5000+ tokens) and LARGE OUTPUT (5000+ tokens)
+This simulates real content generation scenarios
+"""
+import os, sys, django
+sys.path.insert(0, '/app')
+os.environ.setdefault('DJANGO_SETTINGS_MODULE', 'igny8_core.settings')
+django.setup()
+
+import requests
+from igny8_core.modules.system.models import IntegrationSettings
+from igny8_core.auth.models import Account
+
+account = Account.objects.filter(slug='aws-admin').first()
+settings = IntegrationSettings.objects.filter(integration_type='openai', account=account, is_active=True).first()
+config = settings.config or {}
+api_key = config.get('apiKey')
+model = config.get('model', 'gpt-4o')
+
+print(f"🧪 Testing {model} with LARGE INPUT + LARGE OUTPUT")
+print("=" * 80)
+print()
+
+# Create a VERY LONG prompt to ensure input > 5000 tokens
+# Including detailed outline, examples, and extensive instructions
+large_prompt = """You are a professional content writer. Generate a complete, comprehensive article following this detailed specification.
+
+## Article Topic: "Complete Guide to Organic Cotton Bedding: Benefits, Types, and Buying Guide"
+
+## Target Word Count: 2500+ words (approximately 3500+ tokens in output)
+
+## JSON Output Format Required:
+{
+  "title": "article title",
+  "meta_title": "SEO title under 60 chars",
+  "meta_description": "SEO description under 160 chars",
+  "content": "<full HTML content>",
+  "word_count": actual_word_count,
+  "primary_keyword": "organic cotton bedding",
+  "secondary_keywords": ["eco-friendly bedding", "sustainable sheets", "organic duvet covers"],
+  "tags": ["organic bedding", "sustainable home", "eco-friendly textiles", "cotton sheets", "bedroom essentials"],
+  "categories": ["Home & Living > Bedding", "Sustainable Living > Eco-Friendly Products"]
+}
+
+## DETAILED CONTENT STRUCTURE (Each section must be fully written with examples, data, and depth):
+
+### Section 1: Introduction (300 words)
+Write a compelling introduction that:
+- Opens with an engaging hook about the rise of organic bedding
+- Explains why consumers are switching from conventional to organic cotton
+- Discusses market trends and growth statistics
+- Mentions health concerns with conventional bedding
+- Previews what readers will learn in this comprehensive guide
+- Naturally incorporates the primary keyword "organic cotton bedding"
+
+### Section 2: Understanding Organic Cotton (400 words)
+Write a detailed section covering:
+- Definition of organic cotton and GOTS (Global Organic Textile Standard) certification
+- Differences between organic and conventional cotton farming practices
+- Specific pesticides and chemicals avoided in organic farming
+- Water usage statistics (organic vs conventional)
+- Soil health and crop rotation practices
+- Carbon footprint comparison data
+- Other certifications to look for (OEKO-TEX, Fair Trade, etc.)
+- Include specific examples of organic cotton farms and their practices
+- Mention leading organic cotton producing regions globally
+
+### Section 3: Health Benefits of Organic Cotton Bedding (400 words)
+Comprehensive coverage of:
+- Hypoallergenic properties and why they matter
+- Impact on sensitive skin and skin conditions (eczema, psoriasis, allergies)
+- Chemical residues in conventional bedding and health risks
+- Dermatologist recommendations and clinical studies
+- Respiratory benefits from chemical-free textiles
+- Benefits for babies and children
+- Real customer testimonials and experiences
+- Comparison of skin irritation rates between organic and conventional bedding
+- Long-term health advantages
+
+### Section 4: Environmental Impact and Sustainability (400 words)
+Detailed analysis including:
+- Water conservation statistics (gallons saved per pound of cotton)
+- Pesticide elimination impact on ecosystems
+- Biodiversity benefits for pollinators and wildlife
+- Soil degradation prevention through organic farming
+- Carbon emissions comparison with conventional cotton
+- Impact on farmer health and community welfare
+- Waste reduction in textile production
+- End-of-life biodegradability
+- Circular economy considerations
+- Case studies of sustainable bedding brands
+
+### Section 5: Types of Organic Cotton Bedding (350 words)
+Comprehensive overview of:
+- Organic cotton sheets (fitted, flat, pillowcases)
+  - Thread count ranges and what they mean
+  - Weave types: percale, sateen, jersey, flannel
+  - Specific characteristics of each weave
+- Organic cotton duvet covers and comforters
+  - Fill power and warmth ratings
+  - Seasonal considerations
+- Organic cotton mattress protectors and pads
+  - Waterproof vs breathable options
+- Organic cotton blankets and throws
+- Specialty items (body pillows, Euro shams, bed skirts)
+- Price ranges for each category
+- Brand examples for each product type
+
+### Section 6: Quality Factors and What to Look For (350 words)
+In-depth buying guidance:
+- Thread count myths and realities (why 800+ isn't always better)
+- Ply differences (single-ply vs multi-ply)
+- Weave construction details
+- Fabric weight and GSM (grams per square meter)
+- Fiber length and quality grades
+- Color fastness and dye methods (low-impact dyes)
+- Shrinkage rates and pre-washing
+- Durability testing standards
+- Pilling resistance
+- How to verify authentic organic certification
+- Red flags for fake "organic" claims
+
+### Section 7: Top Organic Cotton Bedding Brands (400 words)
+Detailed brand reviews including:
+- Coyuchi: Product range, price points, unique features, customer ratings
+- Boll & Branch: Signature products, fair trade practices, pricing
+- Parachute: Material quality, design aesthetic, value proposition
+- Pottery Barn Organic: Accessibility, variety, retail presence
+- West Elm Organic: Style options, price range, sustainability initiatives
+- SOL Organics: Budget-friendly options, certification details
+- Avocado: Premium positioning, additional eco-features
+- Include specific product examples from each brand
+- Price comparison table format
+- Warranty and return policy highlights
+
+### Section 8: Care and Maintenance (300 words)
+Comprehensive care instructions:
+- Initial washing recommendations before first use
+- Water temperature guidelines for different weaves
+- Detergent recommendations (eco-friendly, fragrance-free)
+- Drying methods: air-dry vs machine dry
+- Ironing tips for different weaves
+- Stain removal techniques for organic cotton
+- Frequency of washing for longevity
+- Storage recommendations for seasonal items
+- How to maintain whiteness without bleach
+- Expected lifespan with proper care (5-10 years)
+- When to replace bedding
+
+### Section 9: Cost Analysis and Value Proposition (300 words)
+Financial breakdown:
+- Price range overview: budget ($50-100), mid-range ($100-250), premium ($250-500+)
+- Cost per use calculation over 5 years
+- Durability comparison with conventional bedding
+- Health cost savings (reduced allergies, better sleep)
+- Environmental value beyond personal use
+- Sale and discount timing (when to buy)
+- Starter set recommendations for budget-conscious buyers
+- Investment prioritization (which pieces to buy first)
+- Comparison with luxury conventional bedding prices
+
+### Section 10: Where to Buy (200 words)
+Shopping guide including:
+- Direct-to-consumer brands (online)
+- Major retailers (Pottery Barn, West Elm, Crate & Barrel)
+- Amazon and online marketplaces (verification tips)
+- Local organic stores and boutiques
+- Outlet and discount sources
+- International shipping considerations
+- Return policies and trial periods
+- Customer service quality comparisons
+
+### Section 11: Conclusion (150 words)
+Wrap up with:
+- Summary of key benefits
+- Final recommendations for different buyer types
+- Call to action
+- Future of organic bedding industry
+
+## Writing Requirements:
+- Use active voice throughout
+- Include specific data points, statistics, and examples
+- Vary sentence structure for natural flow
+- Use HTML tags properly: <p>, <h2>, <h3>, <ul>, <ol>, <table>
+- Create comparison tables where appropriate
+- Use <blockquote> for expert quotes or key insights
+- Maintain professional, editorial tone
+- Avoid generic phrases and filler content
+- Each paragraph should be 3-5 sentences minimum
+- Lists should have detailed descriptions, not just bullet points
+
+Return ONLY the JSON object. The content field must contain ALL sections fully written in HTML format."""
+
+print("📊 Prompt Statistics:")
+print(f"   Prompt Length: {len(large_prompt)} characters")
+print(f"   Estimated Input Tokens: ~{len(large_prompt) // 4} tokens")
+print()
+
+# Test with different max_tokens settings
+tests = [
+    {"name": "Test 1: max_tokens=4096", "max_tokens": 4096},
+    {"name": "Test 2: max_tokens=8192", "max_tokens": 8192},
+    {"name": "Test 3: max_tokens=16384", "max_tokens": 16384},
+]
+
+print("🚀 Starting API Tests...")
+print("=" * 80)
+
+for test in tests:
+    print(f"\n{test['name']}")
+    print("-" * 80)
+    
+    try:
+        response = requests.post(
+            'https://api.openai.com/v1/chat/completions',
+            headers={
+                'Authorization': f'Bearer {api_key}',
+                'Content-Type': 'application/json',
+            },
+            json={
+                'model': model,
+                'messages': [{'role': 'user', 'content': large_prompt}],
+                'max_tokens': test['max_tokens'],
+                'temperature': 0.7,
+            },
+            timeout=180  # 3 minutes for long generation
+        )
+        
+        if response.status_code == 200:
+            data = response.json()
+            usage = data.get('usage', {})
+            finish_reason = data['choices'][0].get('finish_reason')
+            
+            prompt_tokens = usage.get('prompt_tokens', 0)
+            completion_tokens = usage.get('completion_tokens', 0)
+            total_tokens = usage.get('total_tokens', 0)
+            
+            # Estimate word count (1 token ≈ 0.75 words)
+            estimated_words = int(completion_tokens * 0.75)
+            
+            print(f"✓ Response received")
+            print(f"   Input Tokens:      {prompt_tokens:>6,} tokens")
+            print(f"   Output Tokens:     {completion_tokens:>6,} tokens")
+            print(f"   Total Tokens:      {total_tokens:>6,} tokens")
+            print(f"   Estimated Words:   ~{estimated_words:>5,} words")
+            print(f"   Finish Reason:     {finish_reason}")
+            
+            if finish_reason == 'length':
+                print(f"   🚨 TRUNCATED: Hit max_tokens={test['max_tokens']:,} limit!")
+                print(f"   ⚠️  Content is INCOMPLETE - needs higher max_tokens")
+            elif finish_reason == 'stop':
+                print(f"   ✅ COMPLETE: Response finished naturally")
+            
+            # Check if we met our goals
+            if prompt_tokens > 5000:
+                print(f"   ✅ Input > 5000 tokens")
+            else:
+                print(f"   ⚠️  Input only {prompt_tokens} tokens (target: 5000+)")
+            
+            if completion_tokens > 5000:
+                print(f"   ✅ Output > 5000 tokens")
+            else:
+                print(f"   ⚠️  Output only {completion_tokens} tokens (target: 5000+)")
+                
+        else:
+            error_data = response.json()
+            error_msg = error_data.get('error', {}).get('message', 'Unknown error')
+            print(f"   ✗ API Error: {response.status_code}")
+            print(f"   Error: {error_msg}")
+    
+    except requests.exceptions.Timeout:
+        print(f"   ✗ Request timed out after 180 seconds")
+    except Exception as e:
+        print(f"   ✗ Exception: {str(e)}")
+
+print("\n" + "=" * 80)
+print("🎯 FINAL ANALYSIS")
+print("=" * 80)
+print()
+print("Key Questions Answered:")
+print()
+print("1. Can the model handle 5000+ token INPUT?")
+print("   → Check 'Input Tokens' above (should be 5000+)")
+print()
+print("2. Can the model generate 5000+ token OUTPUT?")
+print("   → Check 'Output Tokens' above (should be 5000+)")
+print()
+print("3. Does max_tokens limit actually work as expected?")
+print("   → Compare output tokens across tests with different limits")
+print()
+print("4. Is content getting truncated with current 8192 setting?")
+print("   → Check if 'Finish Reason' is 'length' for 8192 test")
+print()
+print("💡 RECOMMENDATION:")
+print("   If Test 2 (8192) shows 'TRUNCATED', increase max_tokens for content generation")
+print("   If Test 3 (16384) shows 'COMPLETE', that's the minimum needed for long articles")
+print()