Image Generation and Analysis with DIAL API

A Python implementation task to work with AI image generation and analysis capabilities via DIAL API

🎓 Learning Goals

By completing these tasks, you will learn:

• How to generate images from text prompts using DALL-E 3
• How to analyze images using different AI models (GPT-4o, Claude-3-Sonnet)
• Two different approaches for handling images in AI systems:
  • OpenAI approach: Base64 encoding for direct embedding
  • DIAL approach: Bucket storage with attachment references
• How to work with file uploads, downloads with DIAL bucket
• How DIAL adapts requests for different AI model vendors

---

📋 Requirements

• Python 3.11+
• pip
• API key for DIAL service

🔧 Setup

1. Install dependencies:

   pip install -r requirements.txt

2. Set your API key:

   • Ensure that you are connected to the EPAM VPN
   • Get the DIAL API key here: https://support.epam.com/ess?id=sc_cat_item&table=sc_cat_item&sys_id=910603f1c3789e907509583bb001310c
   • Update the API_KEY constant in task/_utils/constants.py
   • Get available models from: https://ai-proxy.lab.epam.com/openai/models

3. Project structure:

   task/
   ├── _models/
   │   ├── conversation.py          # ✅ Complete
   │   ├── message.py               # ✅ Complete  
   │   ├── role.py                  # ✅ Complete
   │   └── custom_content.py        # ✅ Complete
   ├── _utils/
   │   ├── model_client.py          # ✅ Complete
   │   ├── bucket_client.py         # ✅ Complete
   │   ├── constants.py             # ✅ Complete
   │   └── request.py               # ✅ Complete
   ├── image_to_text/
   │   ├── openai/
   │   │   ├── message.py           # ✅ Complete
   │   │   └── task_openai_itt.py   # 🚧 TODO
   │   └── task_dial_itt.py         # 🚧 TODO
   └── text_to_image/
       └── task_tti.py              # 🚧 TODO
   dialx-banner.png                 # 📁 Sample image
   

---

📝 Your Tasks

If the task in the main branch is hard for you, then switch to the with-detailed-description branch

Complete the implementation of these three practice files:

1. task_openai_itt.py - OpenAI-Style Image Analysis

Goal: Analyze an image using base64 encoding approach

• Create DialModelClient with GPT-4o model (and other models)
• Encode image as base64 data URL
• Send ContentedMessage with text and image content
• Key Learning: Direct image embedding in messages

2. task_dial_itt.py - DIAL-Style Image Analysis

Goal: Analyze an image using bucket storage approach

• Upload image to DIAL bucket storage
• Create message with attachment reference
• Test with different AI models
• Key Learning: File storage and attachment handling

3. task_tti.py - Text-to-Image Generation

Goal: Generate images from text prompts

• Create text prompt for image generation
• Use DALL-E 3 model for generation
• Download and save generated images
• Experiment with size, quality and style of output via custom_fields configuration parameter
• Key Learning: AI image generation and file handling

🎯 Expected Outputs

Text-to-Image Task

• Generated image file saved locally with timestamp
• Console output showing request/response details

Image-to-Text Tasks

• AI description of the dialx-banner.png image
• Comparison between different models' responses

🛠️ Troubleshooting

Common Issues:

• Empty API key: Update constants.py with your DIAL API key
• VPN connection: Ensure EPAM VPN is active
• File not found: Verify dialx-banner.png exists in project root
• Network errors: Check DIAL service status and connectivity

🌟 Bonus Challenges

Once you complete the basic tasks, try these extensions:

1. Multi-model comparison: Run the same image analysis with different models
2. Custom prompts: Create your own text-to-image prompts
3. Batch processing: Analyze multiple images at once
4. Error handling: Add robust error handling and retries

📚 Key Concepts Covered

• Multimodal AI: Working with both text and images
• API Design Patterns: Different approaches to handle media content
• Async Programming: File operations and HTTP requests
• Model Abstraction: How DIAL Core adapts requests across vendors
• File Management: Upload, download, and storage operations

---