Skip to content

[5/n] DP Enhancement: Correct num_token_non_padded #9107

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 3 commits into from
Aug 12, 2025
Merged

[5/n] DP Enhancement: Correct num_token_non_padded #9107

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
881b6e1
fix num_tokens_non_padded
ch-wan Aug 12, 2025
659f888
fix
ch-wan Aug 12, 2025
b488698
Merge branch 'main' into cheng/dev/correct-num-non-padded-token
ch-wan Aug 12, 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
21 changes: 19 additions & 2 deletions python/sglang/srt/model_executor/cuda_graph_runner.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -33,7 +33,11 @@
set_graph_pool_id,
)
from sglang.srt.distributed.parallel_state import GroupCoordinator, graph_capture
from sglang.srt.layers.dp_attention import DPPaddingMode, get_attention_tp_size
from sglang.srt.layers.dp_attention import (
DPPaddingMode,
get_attention_tp_rank,
get_attention_tp_size,
)
from sglang.srt.layers.logits_processor import LogitsProcessorOutput
from sglang.srt.layers.torchao_utils import save_gemlite_cache
from sglang.srt.model_executor.forward_batch_info import (
Expand Down Expand Up @@ -255,6 +259,9 @@ def __init__(self, model_runner: ModelRunner):
self.dp_size = model_runner.server_args.dp_size
self.pp_size = model_runner.server_args.pp_size

self.attn_tp_size = get_attention_tp_size()
self.attn_tp_rank = get_attention_tp_rank()

# Batch sizes to capture
self.capture_bs, self.compile_bs = get_batch_sizes_to_capture(model_runner)
rank0_log(f"Capture cuda graph bs {self.capture_bs}")
Expand Down Expand Up @@ -749,7 +756,17 @@ def replay_prepare(
self.global_num_tokens_gpu.fill_(bs * self.num_tokens_per_bs)
self.global_num_tokens_for_logprob_gpu.fill_(bs * self.num_tokens_per_bs)
if enable_num_token_non_padded(self.model_runner.server_args):
self.num_token_non_padded.copy_(forward_batch.num_token_non_padded)
num_token_non_padded = forward_batch.num_token_non_padded
if self.require_gathered_buffer:
tokens_per_rank = bs // self.attn_tp_size * self.num_tokens_per_bs
num_local_token_non_padded = torch.clamp(
num_token_non_padded - tokens_per_rank * self.attn_tp_rank,
min=0,
max=tokens_per_rank,
)
self.num_token_non_padded.copy_(num_local_token_non_padded)
else:
self.num_token_non_padded.copy_(num_token_non_padded)
if self.enable_two_batch_overlap:
self.tbo_plugin.replay_prepare(
forward_mode=self.capture_forward_mode,
Expand Down
Loading