Skip to content

update dlblas with deepep #4126

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

Open
wants to merge 4 commits into
base: main
Choose a base branch
from
Open

update dlblas with deepep #4126

Changes from 1 commit
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
febfa9e
update_dlblas_with_deepep
SHshenhao Nov 14, 2025
c8ce820
update_dlblasanddeepep
SHshenhao Nov 18, 2025
6cc9e6f
Merge remote-tracking branch 'origin/main' into sh/update_dlblas
SHshenhao Nov 18, 2025
6a7c734
update_dlblas_deepep
SHshenhao Nov 18, 2025
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
6 changes: 6 additions & 0 deletions lmdeploy/pytorch/models/deepseek_v2.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1169,6 +1169,12 @@ def prepare_inputs_for_generation(
position_ids = context.position_ids
attn_metadata = context.attn_metadata

from dlblas.layers.moe.token_dispatcher import DeepEPBuffer, DeepEPMode
Copy link
Collaborator

@grimoire grimoire Nov 16, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

move this to backend

lmdeploy/lmdeploy/pytorch/backends/cuda/moe.py

Line 450 in f63730d

class FusedDeepEpMoEBlockedF8Impl(TritonFusedMoEBlockedF8Impl):

You can get context with context manager

lmdeploy/lmdeploy/pytorch/model_inputs.py

Line 531 in f63730d

def get_step_ctx_manager():

deepep_mode = DeepEPMode.NORMAL
if context.is_decoding:
deepep_mode = DeepEPMode.LOW_LATENCY
DeepEPBuffer.set_deepep_mode(deepep_mode)

return dict(
input_ids=input_ids,
position_ids=position_ids,
Expand Down
Loading