Skip to content

fix missing enable_pdl argument in trtllm-gen fp4 moe #1480

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 4 commits into from
Aug 14, 2025
Merged

fix missing enable_pdl argument in trtllm-gen fp4 moe #1480

Changes from 1 commit
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
146b985
fix enable_pdl
IwakuraRein Aug 13, 2025
c7d351f
Update flashinfer/fused_moe/core.py
IwakuraRein Aug 14, 2025
9558745
address pre-commit
IwakuraRein Aug 14, 2025
88538f1
upd
yyihuang Aug 14, 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
7 changes: 7 additions & 0 deletions flashinfer/fused_moe/core.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -593,6 +593,7 @@ def _fake_cutlass_fused_moe_sm100(
use_mxfp8_act_scaling: bool = False,
min_latency_mode: bool = False,
tune_max_num_tokens: int = 8192,
enable_pdl: Optional[bool] = None,
):
seq_len = input.shape[0]
hidden_size = fc2_expert_weights.shape[1]
Expand Down Expand Up @@ -947,6 +948,7 @@ def _fake_trtllm_fp8_per_tensor_scale_moe(
use_routing_scales_on_input: bool,
tile_tokens_dim: int = 8,
routing_method_type: int = 0,
enable_pdl: Optional[bool] = None,
):
seq_len = hidden_states.shape[0]
hidden_size = hidden_states.shape[1]
Expand Down Expand Up @@ -1031,6 +1033,7 @@ def _fake_trtllm_fp8_block_scale_moe(
routing_method_type: int = 0,
use_shuffled_weight: bool = False,
weight_layout: int = 0,
enable_pdl: Optional[bool] = None,
):
seq_len = hidden_states.shape[0]
hidden_size = hidden_states.shape[1]
Expand Down Expand Up @@ -1174,6 +1177,8 @@ def _fake_trtllm_fp4_block_scale_moe(
tile_tokens_dim: int,
routing_method_type: int,
do_finalize: bool,
enable_pdl: Optional[bool],
output: Optional[torch.Tensor],
):
seq_len = hidden_states.shape[0]
hidden_size = hidden_states.shape[1]
Expand Down Expand Up @@ -1360,6 +1365,7 @@ def trtllm_fp4_block_scale_moe(
tile_tokens_dim: int = 8,
routing_method_type: int = 0,
do_finalize: bool = True,
enable_pdl: Optional[bool] = None,
output: Optional[torch.Tensor] = None,
) -> List[torch.Tensor]:
"""FP4 block scale MoE operation.
Expand Down Expand Up @@ -1440,6 +1446,7 @@ def trtllm_fp4_block_scale_moe(
tile_tokens_dim,
routing_method_type,
do_finalize,
enable_pdl,
output,
)

Expand Down