Skip to content

[BUG] fix local_rank in initialize_dp_attention #7584

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
zhyncs merged 4 commits into from
Jun 28, 2025
Merged

[BUG] fix local_rank in initialize_dp_attention #7584

Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
13edfb8
fix local_rank in initialize_dp_attention
TomQuartz Jun 27, 2025
b9e6151
Merge branch 'main' into fix/dp_attention
ch-wan Jun 27, 2025
d2eabf7
rm comments; directly use tp_group.local_rank
TomQuartz Jun 28, 2025
a869541
Merge branch 'main' into fix/dp_attention
zhyncs Jun 28, 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
4 changes: 1 addition & 3 deletions python/sglang/srt/layers/dp_attention.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -79,14 +79,12 @@ def initialize_dp_attention(
)

if enable_dp_attention:
local_rank = tp_rank % (tp_size // dp_size)
_ATTN_DP_SIZE = dp_size
if moe_dense_tp_size is None:
_LOCAL_ATTN_DP_SIZE = _ATTN_DP_SIZE
else:
_LOCAL_ATTN_DP_SIZE = max(1, dp_size // (tp_size // moe_dense_tp_size))
else:
local_rank = tp_rank
_ATTN_DP_SIZE = 1
_LOCAL_ATTN_DP_SIZE = 1

Expand All @@ -96,7 +94,7 @@ def initialize_dp_attention(
list(range(head, head + _ATTN_TP_SIZE))
for head in range(0, pp_size * tp_size, _ATTN_TP_SIZE)
],
local_rank,
tp_group.local_rank,
torch.distributed.get_backend(tp_group.device_group),
use_pynccl=SYNC_TOKEN_IDS_ACROSS_TP,
use_pymscclpp=False,
Expand Down
Loading