NousResearch/atropos
1
0
Fork 0
mirror of https://github.com/NousResearch/atropos.git synced 2026-04-19 12:57:58 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

correction

Browse source
This commit is contained in:
Jai Suphavadeeprasit 2025-12-03 16:44:17 -05:00
parent 61221dd1a2
commit 2d3c07dcae
1 changed files with 6 additions and 6 deletions
Download patch file Download diff file Expand all files Collapse all files

12
example_trainer/vllm_api_server.py
View file

@ -21,7 +21,7 @@ Architecture:
│ └───────┼────────────┼────────────────┼──────────┘ │
│ │ │ │ │
│ ┌───────▼────────────▼────────────────▼──────────┐ │
│ │ AsyncLLMEngine │ │
│ │ AsyncLLM │ │
│ │ - Model weights (shared via NCCL) │ │
│ │ - LoRA adapters (hot-swappable) │ │
│ └────────────────────────────────────────────────┘ │
@ -52,7 +52,7 @@ from vllm.logger import init_logger
from vllm.sampling_params import RequestOutputKind, SamplingParams
from vllm.usage.usage_lib import UsageContext
from vllm.utils import random_uuid
from vllm.v1.engine.async_llm import AsyncLLMEngine
from vllm.v1.engine.async_llm import AsyncLLM
try:
from vllm.utils.argparse_utils import FlexibleArgumentParser
@ -70,7 +70,7 @@ logger = init_logger("vllm.entrypoints.api_server")
# =============================================================================
app = FastAPI()
engine: Optional[AsyncLLMEngine] = None
engine: Optional[AsyncLLM] = None
@dataclass
@ -494,7 +494,7 @@ def build_app(args: Namespace) -> FastAPI:
async def init_app(
args: Namespace,
llm_engine: AsyncLLMEngine | None = None,
llm_engine: AsyncLLM | None = None,
) -> FastAPI:
"""
Initialize the application and vLLM engine.
@ -513,7 +513,7 @@ async def init_app(
engine = (
llm_engine
if llm_engine is not None
else AsyncLLMEngine.from_engine_args(
else AsyncLLM.from_engine_args(
engine_args, usage_context=UsageContext.API_SERVER
)
)
@ -555,7 +555,7 @@ def _export_state_dict_info(args: Namespace) -> None:
async def run_server(
args: Namespace, llm_engine: AsyncLLMEngine | None = None, **uvicorn_kwargs: Any
args: Namespace, llm_engine: AsyncLLM | None = None, **uvicorn_kwargs: Any
) -> None:
"""
Run the vLLM API server.