iHore/vLLM
8
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

[XPU] fix fused_input_norm for MM (#59865)

Browse Source 
Signed-off-by: Yan Ma <yan.ma@intel.com>
This commit is contained in:
Yan Ma
2026-10-09 20:16:35 +08:00
committed by GitHub
parent 0cc0460870
commit c17fadcfdc
3 changed files with 11 additions and 23 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.buildkite/intel_jobs/model_executor_intel.yaml
+1 -10
View File
@@ -28,16 +28,7 @@ steps:
export VLLM_WORKER_MULTIPROC_METHOD=spawn &&
export PYTHONFAULTHANDLER=1 &&
cd tests &&
pytest -v -s model_executor -m "not slow_test" --ignore="model_executor/layers/test_rocm_unquantized_gemm.py" --ignore="model_executor/model_loader/tensorizer_loader/test_tensorizer.py" --deselect="tests/model_executor/model_loader/test_reload.py::test_kv_scale_reload"
--deselect="tests/model_executor/layers/test_mm_input_norm.py::TestFusedMMInputNormModule::test_matches_reference_above_cudnn_grid_limit"
--deselect="tests/model_executor/layers/test_mm_input_norm.py::TestFusedMMInputNormDtypes::test_dtype_combinations[in_dtype0-out_dtype0]"
--deselect="tests/model_executor/layers/test_mm_input_norm.py::TestFusedMMInputNormDtypes::test_dtype_combinations[in_dtype1-out_dtype1]"
--deselect="tests/model_executor/layers/test_mm_input_norm.py::TestFusedMMInputNormDtypes::test_dtype_combinations[in_dtype2-out_dtype2]"
--deselect="tests/model_executor/layers/test_mm_input_norm.py::TestFusedMMInputNormDtypes::test_dtype_combinations[in_dtype3-out_dtype3]"
--deselect="tests/model_executor/layers/test_mm_input_norm.py::TestFusedMMInputNormDtypes::test_dtype_combinations[in_dtype4-out_dtype4]"
--deselect="tests/model_executor/layers/test_mm_input_norm.py::TestFusedMMInputNormDtypes::test_dtype_combinations[in_dtype5-out_dtype5]"
--deselect="tests/model_executor/layers/test_mm_input_norm.py::TestFusedMMInputNormShapes"
--deselect="tests/model_executor/layers/test_mm_input_norm.py::TestFusedMMInputNormInputHandling"'
pytest -v -s model_executor -m "not slow_test" --ignore="model_executor/layers/test_rocm_unquantized_gemm.py" --ignore="model_executor/model_loader/tensorizer_loader/test_tensorizer.py" --deselect="tests/model_executor/model_loader/test_reload.py::test_kv_scale_reload"'
- label: Tensorizer Model Loader Tests (XPU)
key: tensorizer-model-loader-tests-xpu
.buildkite/intel_jobs/models_multimodal_intel.yaml
+2 -1
View File
@@ -73,7 +73,8 @@ steps:
bash .buildkite/scripts/hardware_ci/run-intel-test.sh
'cd tests &&
pytest -v -s models/multimodal/generation/test_common.py -m core_model -k "not qwen2 and not qwen3 and not gemma" &&
pytest -v -s models/multimodal/generation/test_gemma_ragged_audio.py -m core_model'
pytest -v -s models/multimodal/generation/test_gemma_ragged_audio.py -m core_model
pytest -v -s models/multimodal/generation/test_qwen2_vl.py -m core_model'
- label: "Multi-Modal Models (Standard) 4: other + whisper"
key: multi-modal-models-standard-4-other-whisper
vllm/model_executor/layers/fusion/mm_input_norm.py
+8 -12
View File
@@ -313,19 +313,15 @@ class FusedMMInputNorm(CustomOp):
def forward_xpu(
self, pixel_values: torch.Tensor, visual_dtype: torch.dtype
) -> torch.Tensor:
"""XPU fused custom kernel path.
On XPU, fuse the whole rescale + normalise into a single custom
kernel. The eager path materializes an fp32 intermediate and then
casts back, which adds device-side compute that cancels the
bandwidth saving of transferring uint8 pixel_values. The fused
kernel reads uint8 directly and writes ``visual_dtype`` in one pass.
"""XPU fused custom kernel path for uint8 and Triton kernel path
for others.
"""
# The out-of-tree XPU kernel only supports the uint8 input that
# device-side normalisation guarantees; fail loudly otherwise.
assert pixel_values.dtype == torch.uint8, (
f"xpu_fused_input_norm requires uint8 input, got {pixel_values.dtype}"
)
if pixel_values.dtype != torch.uint8:
return self.forward_cuda(pixel_values, visual_dtype)
# For uint8 input, falls to xpu_fused_input_norm from xpu kernels
import vllm._xpu_ops # noqa: F401
return torch.ops.vllm.xpu_fused_input_norm(
pixel_values, self.weight, self.bias, visual_dtype
)