followup fix for llama 4 trtllm flashinfer backend (#12314)

2025-10-28 22:17:08 -07:00
parent 0380ca82ef
commit 83a7c89c3f
1 changed files with 1 additions and 1 deletions
--- a/python/sglang/srt/layers/quantization/modelopt_quant.py
+++ b/python/sglang/srt/layers/quantization/modelopt_quant.py
@@ -673,7 +673,7 @@ class ModelOptFp8MoEMethod(FusedMoEMethodBase):
                    routed_scaling_factor if routed_scaling_factor is not None else 1.0
                ),
                use_routing_scales_on_input=use_routing_scales_on_input,
-                tile_tokens_dim=None,
+                tile_tokens_dim=8,  # TODO(brayden): use the FI tile calculation
                routing_method_type=routing_method_type,
            )