From 4975fda88247eae37c665deb4fb106e1790548c2 Mon Sep 17 00:00:00 2001 From: Wei Zhao <51183510+wzhao18@users.noreply.github.com> Date: Wed, 14 Oct 2026 15:31:08 -0400 Subject: [PATCH] [Dependency] Upgrade FlashInfer to 0.7.1 (#60948) Signed-off-by: wzhao18 Co-authored-by: Jee Jee Li --- requirements/cuda.txt | 2 +- vllm/models/kimi_k3/nvidia/kda.py | 1 + 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/requirements/cuda.txt b/requirements/cuda.txt index def6f5c4b5..e1d0bc932b 100644 --- a/requirements/cuda.txt +++ b/requirements/cuda.txt @@ -10,7 +10,7 @@ torchaudio==2.11.0 torchvision==0.29.0 # Required for phi3v processor. See https://github.com/pytorch/vision?tab=readme-ov-file#installation for corresponding version torchcodec >= 0.14 PyNvVideoCodec==2.0.4 -flashinfer-python==0.7.0.post1 +flashinfer-python==0.7.1 apache-tvm-ffi==0.1.11 tilelang==0.1.12 nvidia-cudnn-frontend>=1.19.1 diff --git a/vllm/models/kimi_k3/nvidia/kda.py b/vllm/models/kimi_k3/nvidia/kda.py index 26fe7cb5b4..284e4585c6 100644 --- a/vllm/models/kimi_k3/nvidia/kda.py +++ b/vllm/models/kimi_k3/nvidia/kda.py @@ -456,6 +456,7 @@ def _flashinfer_kda_prefill( beta_is_logit=True, seq_order=seq_order, prefill_workspace=prefill_workspace, + backend="cute-dsl", ) return output, initial_state