Fix attention backend logic for Qwen3-Next on SM100 (#14560)

2025-12-06 22:03:34 -08:00
parent 0e4d8790d7
commit 3c7886ec4c
1 changed files with 15 additions and 0 deletions
@@ -1282,6 +1282,21 @@ class ServerArgs:
                    logger.info(
                        "Use flashinfer_trtllm as MoE runner backend on sm100 for Qwen3NextForCausalLM"
                    )
                if self.attention_backend is None:
                    self.attention_backend = "triton"
                    logger.info(
                        "Use triton as attention backend on sm100 for Qwen3NextForCausalLM"
                    )
                if (
                    not self.disable_radix_cache
                    and self.attention_backend == "trtllm_mha"
                ):
                    logger.warning(
                        "Disabling radix cache since trtllm_mha does not support page_size = 1, which is required by MambaRadixCache. "
                        "Try to use --attention-backend triton if radix cache is necessary."
                    )
                    self.disable_radix_cache = True
                    self.disable_overlap_schedule = False
        elif model_arch in [
            "NemotronHForCausalLM",
            "FalconH1ForCausalLM",