Skip to content

Commit e811123

Browse files
BBuffinger92
authored andcommitted
[ci] fix llama4 ci error (sgl-project#5126)
1 parent 7e305c3 commit e811123

File tree

1 file changed

+2
-0
lines changed

1 file changed

+2
-0
lines changed

python/sglang/srt/layers/quantization/compressed_tensors/compressed_tensors_moe.py

Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -285,6 +285,7 @@ def apply(
285285
activation: str = "silu",
286286
inplace: bool = True,
287287
no_combine: bool = False,
288+
apply_router_weight_on_input: bool = False,
288289
) -> torch.Tensor:
289290
from sglang.srt.layers.moe.fused_moe_triton import fused_experts
290291
from sglang.srt.layers.moe.topk import select_experts
@@ -314,6 +315,7 @@ def apply(
314315
w2_scale=layer.w2_weight_scale,
315316
a1_scale=layer.w13_input_scale,
316317
a2_scale=layer.w2_input_scale,
318+
apply_router_weight_on_input=apply_router_weight_on_input,
317319
)
318320

319321

0 commit comments

Comments
 (0)