Skip to content

[None][chore] local imports for AutoDeploy in serve and bench #9199

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Nov 18, 2025
Merged

[None][chore] local imports for AutoDeploy in serve and bench #9199

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 2 additions & 1 deletion tensorrt_llm/bench/benchmark/__init__.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,7 +6,6 @@

from tensorrt_llm import LLM as PyTorchLLM
from tensorrt_llm._tensorrt_engine import LLM
from tensorrt_llm._torch.auto_deploy import LLM as AutoDeployLLM
from tensorrt_llm.bench.benchmark.utils.processes import IterationWriter
from tensorrt_llm.bench.build.build import get_model_config
from tensorrt_llm.bench.dataclasses.configuration import RuntimeConfig
Expand Down Expand Up @@ -115,6 +114,8 @@ def get_llm(runtime_config: RuntimeConfig, kwargs: dict):
kwargs["enable_iter_perf_stats"] = True

elif runtime_config.backend == "_autodeploy":
from tensorrt_llm._torch.auto_deploy import LLM as AutoDeployLLM

kwargs["world_size"] = kwargs.pop("tensor_parallel_size", None)
llm_cls = AutoDeployLLM

Expand Down
3 changes: 2 additions & 1 deletion tensorrt_llm/commands/serve.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,7 +16,6 @@
from tensorrt_llm import LLM as PyTorchLLM
from tensorrt_llm import MultimodalEncoder
from tensorrt_llm._tensorrt_engine import LLM
from tensorrt_llm._torch.auto_deploy.llm import LLM as AutoDeployLLM
from tensorrt_llm._utils import mpi_rank
from tensorrt_llm.executor.utils import LlmLauncherEnvs
from tensorrt_llm.inputs.multimodal import MultimodalServerConfig
Expand Down Expand Up @@ -162,6 +161,8 @@ def launch_server(
if backend == 'pytorch':
llm = PyTorchLLM(**llm_args)
elif backend == '_autodeploy':
from tensorrt_llm._torch.auto_deploy import LLM as AutoDeployLLM

# AutoDeploy does not support build_config
llm_args.pop("build_config", None)
llm = AutoDeployLLM(**llm_args)
Expand Down
Loading