Skip to content

Support dynamic lora updating #472

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
hiyuchang merged 1 commit into from
Jan 9, 2026
+5 −0
Merged

Support dynamic lora updating #472

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
2 changes: 2 additions & 0 deletions tests/common/vllm_test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1228,6 +1228,8 @@ def setUp(self):
self.config.explorer.rollout_model.tensor_parallel_size = 1
self.config.explorer.rollout_model.chat_template = CHAT_TEMPLATE
self.config.explorer.rollout_model.enable_openai_api = True
self.config.explorer.rollout_model.enable_lora = True
self.config.explorer.rollout_model.enable_runtime_lora_updating = True

self.config.check_and_update()
self.engines, self.auxiliary_engines = create_inference_models(self.config)
Expand Down
1 change: 1 addition & 0 deletions trinity/common/config.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -556,6 +556,7 @@ class InferenceModelConfig:

# ! DO NOT SET, automatically set from model.lora_configs
enable_lora: bool = False
enable_runtime_lora_updating: bool = False
lora_modules: Optional[List[Dict]] = None
lora_kwargs: Optional[dict] = field(default_factory=dict)

Expand Down
2 changes: 2 additions & 0 deletions trinity/common/models/vllm_model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -54,6 +54,8 @@ def __init__(
os.environ["VLLM_ENABLE_V1_MULTIPROCESSING"] = "0"
if self.vllm_version >= parse_version("0.11.0"):
os.environ["VLLM_ALLREDUCE_USE_SYMM_MEM"] = "0"
if self.config.enable_runtime_lora_updating:
os.environ["VLLM_ALLOW_RUNTIME_LORA_UPDATING"] = "1"
if not config.enforce_eager:
# To avoid torch compile conflicts when multiple model are started simultaneously.
# remove this when the following PR is released:
Expand Down