refactor: move omni_llm.py to entrypoints directory

- 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.

examples/basic/simple_usage.py

@@ -7,7 +7,7 @@ with a single AR (Autoregressive) stage.
 """
 
 import asyncio
-from vllm_omni.core.omni_llm import OmniLLM, AsyncOmniLLM
+from vllm_omni.entrypoints.omni_llm import OmniLLM, AsyncOmniLLM
 from vllm_omni.config import create_ar_stage_config
 
 

vllm_omni/__init__.py

@@ -14,7 +14,7 @@ __author__ = "vLLM-omni Team"
 __email__ = "hsliuustc@gmail.com"
 
 # Main entry points
-from .core.omni_llm import OmniLLM, AsyncOmniLLM
+from .entrypoints.omni_llm import OmniLLM, AsyncOmniLLM
 from .config import (
     OmniStageConfig,
     DiTConfig,

vllm_omni/entrypoints/api_server.py

@@ -9,7 +9,7 @@ from fastapi.responses import JSONResponse
 from pydantic import BaseModel
 import uvicorn
 
-from ..core.omni_llm import AsyncOmniLLM
+from .omni_llm import AsyncOmniLLM
 
 
 class GenerateRequest(BaseModel):

vllm_omni/entrypoints/omni.py

@@ -5,7 +5,7 @@ Omni serve command for vLLM-omni.
 import argparse
 import asyncio
 from typing import List, Optional
-from vllm_omni.core.omni_llm import AsyncOmniLLM
+from .omni_llm import AsyncOmniLLM
 from vllm_omni.config import create_ar_stage_config, create_dit_stage_config, DiTConfig
 
 

vllm_omni/entrypoints/omni_llm.py

@@ -10,7 +10,7 @@ from vllm.outputs import RequestOutput, LoRARequest
 
 
 from ..config import OmniStageConfig
-from .stage_manager import StageManager
+from ..core.stage_manager import StageManager
 from ..engine.output_processor import MultimodalOutputProcessor
 
 

vllm_omni/plugin.py

@@ -3,7 +3,7 @@ vLLM plugin system for vLLM-omni integration.
 """
 
 from typing import Dict, Any, Optional
-from vllm_omni.core.omni_llm import OmniLLM, AsyncOmniLLM
+from vllm_omni.entrypoints.omni_llm import OmniLLM, AsyncOmniLLM
 from vllm_omni.config import create_ar_stage_config, create_dit_stage_config