- Introduced OmniGPUModelRunner for enhanced GPU model execution with state management.
- Added OmniModelInputForGPU and its builder to support additional information in model inputs.
- Implemented methods for updating request states and handling multimodal inputs.
- Enhanced input preparation for CUDA graph execution and LoRA integration.
- Introduced `DiffusionModelRunner` class to handle diffusion processes, extending `OmniGPUModelRunner`.
- Updated `DiffusionGPUWorker` to initialize device settings and manage memory more effectively.
- Enhanced model execution with support for multimodal inputs and improved tensor management.
- Removed legacy code and streamlined the generation process for better performance.
- Introduced ARModelRunner for autoregressive GPU model execution, enabling per-request hidden state outputs and token sampling.
- Added ARGPUWorker to initialize GPU settings and manage distributed environment for the AR model runner.
- Enhanced memory management and error handling for GPU resources during initialization.
- Introduced `OmniModelRunnerOutput` and `OmniRequestOutput` classes for handling multimodal outputs.
- Implemented `OmniEngineCoreRequest` and `OmniInputPreprocessor` to manage prompt embeddings and additional information.
- Created `OmniTokensPrompt`, `OmniTokenInputs`, and `OmniEmbedsPrompt` for enhanced token handling.
- Added parsing and preprocessing functions to support new input structures.
- Removed the obsolete `diffusion_engine.py` file as part of the refactor.
- Introduced `DiffusionScheduler` to handle scheduling for diffusion models, allowing immediate request completion.
- Added `OmniNewRequestData` to encapsulate additional request data, including prompt embeddings and extra information.
- Updated `OmniScheduler` to ensure compatibility with new request data structures.
- Created `output.py` to define new data classes for enriched request handling.
- Enhanced `__init__.py` in the engine module to support new payload structures for prompt embeddings and additional information.
- Introduced OmniRequestState to manage multimodal request states.
- Enhanced MultimodalOutputProcessor to handle various output types including images, text, and latents.
- Implemented methods for accumulating multimodal tensors and processing outputs.
- Updated output handling to ensure compatibility with vLLM's base processor while allowing custom modality handlers.
- Add test_serving.sh: Full-featured testing suite with comprehensive validation
- Add quick_test.sh: Fast validation script for quick testing after changes
- Add scripts/README.md: Complete documentation for testing scripts
- Include health checks, text generation, performance testing, and API integration
- Add retry mechanisms and proper error handling
- Support for different models and ports
- Comprehensive logging and colored output
- Ready for CI/CD integration
Usage:
- Quick test: ./scripts/quick_test.sh [port]
- Full test: ./scripts/test_serving.sh [model_path] [port]
- Move vllm_omni/core/omni_llm.py to vllm_omni/entrypoints/omni_llm.py
- Update all import statements across the codebase to reflect new location
- Fix relative imports within the moved file
- Maintain functionality while improving code organization
- All imports and functionality tested and working correctly
This change better reflects that OmniLLM and AsyncOmniLLM are the main entry points
for vLLM-omni functionality, rather than core implementation details.
- Improve API server with better error handling and response formatting
- Enhance CLI with additional options for DiT stages and configuration
- Add comprehensive examples in examples/basic/ including:
- API client with health checks and text generation
- Docker setup and usage examples
- Simple usage patterns for different scenarios
- Add utility scripts for model downloading and Docker setup
- Update documentation with implementation details and testing guidelines
- Fix configuration validation issues in OmniLLM
- Improve stage configuration handling for AR and DiT stages
- Add proper error handling and fallback mechanisms
Tested with Qwen3-0.6B model:
- Server starts successfully on port 8000
- Health and info endpoints working correctly
- Text generation with various parameters functioning
- API client examples working as expected
- CLI help and configuration options working properly
- Add comprehensive PRD, architecture design, and test design documents
- Implement core modules: OmniLLM, AsyncOmniLLM, stage configurations
- Add DiT scheduler and cache manager for diffusion models
- Implement CLI integration with --omni flag support
- Add API server and plugin system for vLLM integration
- Create comprehensive test suite with fixtures
- Update dependencies to vLLM 0.10.2 and PyTorch 2.8.0
- Add conda environment setup and package installation
- Implement stage-based processing architecture
- Add multimodal output processing capabilities
This commit establishes the foundation for multi-modality models
inference and serving with non-autoregressive structures.
- Add complete directory structure for multimodal and non-autoregressive processing
- Include AsyncLLM, OmniEngine, and DiffusionEngine modules
- Add processors, schedulers, executors, and output processors
- Add API interfaces for Higress, Gradio, and ComfyUI
- Add configuration management and CLI interface
- Add comprehensive test structure and documentation
- Add pyproject.toml with dependencies and project configuration
- Add examples and documentation structure
tzhouam