[Feature] Add OpenAI DALL-E compatible image generation API

Builds on @fake0fan's diffusion online serving implementation to provide
a production-ready, OpenAI-compatible image generation API. Implements
the DALL-E /v1/images/generations and /v1/images/edits endpoints with
full async support and proper error handling.

**OpenAI DALL-E API Compatibility:**
- `/v1/images/generations` - Text-to-image generation
- `/v1/images/edits` - Image editing with text prompts
- Full compatibility with OpenAI Python SDK
- Request/response formats match DALL-E specification

**Unified Async Server:**
- Single `vllm serve <model> --omni` command for all diffusion models
- Async AsyncOmniDiffusion engine with thread-pool execution
- Exposes both DALL-E endpoints and /v1/chat/completions
- Automatic model type detection (diffusion vs chat)

**Model Support:**
- Qwen/Qwen-Image (text-to-image with true CFG)
- Tongyi-MAI/Z-Image-Turbo (fast generation, ~9 steps)
- Qwen/Qwen-Image-Edit (image editing)
- Model profiles for per-model defaults and constraints

**Features:**
- Pydantic validation for all request parameters
- Comprehensive error handling with proper HTTP status codes
- Model field validation (request must match server)
- Empty prompt validation
- Response format validation (only b64_json supported)
- File upload validation (size, type, max 4MB)
- Prompt logging at debug level (security-conscious)

**Added:**
- image_api_utils.py - Shared helper functions (parse_size, encode_image, etc.)
- image_model_profiles.py - Model-specific configurations and constraints
- protocol/images.py - Pydantic models for requests/responses
- DALL-E endpoints in api_server.py (/v1/images/generations, /v1/images/edits)
- Comprehensive test suite (38 tests)
- Documentation and example client

**Modified:**
- api_server.py - Integrated DALL-E endpoints with async support
- pipeline_qwen_image_edit.py - Fixed duplicate methods, dimension unpacking, parameter naming

**Start server:**
```bash
vllm serve Qwen/Qwen-Image --omni --port 8000
```

**Generate image:**
```bash
curl -X POST http://localhost:8000/v1/images/generations \
  -H "Content-Type: application/json" \
  -d '{"prompt": "a cat on a laptop", "size": "1024x1024"}'
```

**Edit image:**
```bash
curl -X POST http://localhost:8000/v1/images/edits \
  -F "[email protected]" \
  -F "prompt=make the sky blue"
```

**OpenAI SDK:**
```python
from openai import OpenAI
client = OpenAI(base_url="http://localhost:8000/v1", api_key="none")
response = client.images.generate(
    prompt="a sunset over mountains",
    size="1024x1024"
)
```

Testing available in:

```bash
pytest tests/entrypoints/openai/test_image_server.py -v
```

Built on @fake0fan's excellent diffusion online serving work (commits
e482de0, 003e8f3). This PR adds the DALL-E compatible API layer with
full validation, error handling, and production-ready features.

Signed-off-by: dougbtv <[email protected]>

Author: dougbtv

.buildkite/scripts/simple_test.sh

@@ -50,3 +50,4 @@ VENV_PYTHON="${VENV_DIR}/bin/python"
 "${UV_BIN}" pip install --python "${VENV_PYTHON}" vllm==0.11.0
 "${UV_BIN}" pip install --python "${VENV_PYTHON}" -e ".[dev]"
 "${VENV_PYTHON}" -m pytest -v -s tests/test_omni_llm.py
+"${VENV_PYTHON}" -m pytest -v -s tests/entrypoints/openai/test_image_server.py