xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2025-12-12 02:45:02 +08:00
Code Issues Packages Projects Releases Wiki Activity

[BugFix] Fix plan API Mismatch when using latest FlashInfer (#29426)

Browse Source 
Signed-off-by: Andrii Skliar <askliar@askliar-mlt.client.nvidia.com>
Co-authored-by: Andrii Skliar <askliar@askliar-mlt.client.nvidia.com>
This commit is contained in:
Andrii Skliar 2025-11-27 20:34:59 +01:00 committed by GitHub
parent e5a621b724
commit a5345bf49d
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
3 changed files with 5 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
docker/Dockerfile
View File

@ -398,8 +398,8 @@ RUN --mount=type=bind,from=build,src=/workspace/dist,target=/vllm-workspace/dist
# Install FlashInfer pre-compiled kernel cache and binaries # Install FlashInfer pre-compiled kernel cache and binaries
# https://docs.flashinfer.ai/installation.html # https://docs.flashinfer.ai/installation.html
RUN --mount=type=cache,target=/root/.cache/uv \ RUN --mount=type=cache,target=/root/.cache/uv \
uv pip install --system flashinfer-cubin==0.5.2 \ uv pip install --system flashinfer-cubin==0.5.3 \
&& uv pip install --system flashinfer-jit-cache==0.5.2 \ && uv pip install --system flashinfer-jit-cache==0.5.3 \
--extra-index-url https://flashinfer.ai/whl/cu$(echo $CUDA_VERSION | cut -d. -f1,2 | tr -d '.') \ --extra-index-url https://flashinfer.ai/whl/cu$(echo $CUDA_VERSION | cut -d. -f1,2 | tr -d '.') \
&& flashinfer show-config && flashinfer show-config

2
requirements/cuda.txt
View File

@ -10,4 +10,4 @@ torchaudio==2.9.0
# These must be updated alongside torch # These must be updated alongside torch
torchvision==0.24.0 # Required for phi3v processor. See https://github.com/pytorch/vision?tab=readme-ov-file#installation for corresponding version torchvision==0.24.0 # Required for phi3v processor. See https://github.com/pytorch/vision?tab=readme-ov-file#installation for corresponding version
# FlashInfer should be updated together with the Dockerfile # FlashInfer should be updated together with the Dockerfile
flashinfer-python==0.5.2 flashinfer-python==0.5.3

3
vllm/v1/attention/backends/flashinfer.py
View File

@ -1508,7 +1508,7 @@ def fast_plan_decode(
qo_indptr_host = _get_range_buf(batch_size + 1, "cpu") qo_indptr_host = _get_range_buf(batch_size + 1, "cpu")
try: try:
# Make sure we pass exactly 18 arguments for tensor core version # Make sure we pass exactly 19 arguments for tensor core version
self._plan_info = self._cached_module.plan( self._plan_info = self._cached_module.plan(
self._float_workspace_buffer, self._float_workspace_buffer,
self._int_workspace_buffer, self._int_workspace_buffer,
@ -1528,6 +1528,7 @@ def fast_plan_decode(
window_left, window_left,
fixed_split_size, fixed_split_size,
disable_split_kv, disable_split_kv,
0,
) )
except Exception as e: except Exception as e:
raise RuntimeError(f"Error in tensor core plan: {e}") from e raise RuntimeError(f"Error in tensor core plan: {e}") from e